chore(beads): link curator and memory-engine epics to their children; file RegisterExternal wiring gap
This commit is contained in:
1 parent
e23e9fc989
commit
aefb6c785a
2 files changed
+13
-11
No files matched your search
@@ -155,3 +155,4 @@
|
||||
{"id":"int-659ff8bb1ef5160f8bc3f1e3f31d3d75","kind":"field_change","created_at":"2026-08-31T11:51:41.85053027Z","actor":"Sam Catlow","issue_id":"archie-core-h019.3","extra":{"field":"status","new_value":"closed","old_value":"in_progress","reason":"Implemented as agent.Subagent inside the task-scoped archie-agent worker (internal/app/agentworker/review.go). Conversation isolation is structural (Subagent.Run only sends a prompt, never message history -- verified against a fake chat.Provider). Workspace isolation via new worktree.Manager.Snapshot (.git-free export of HEAD's tracked files into a scratch dir beside the task worktree). Reviewer's toolset is read-only (read/grep/find) rooted at the snapshot, never the worker's own toolset -- verified as a construction fact that the toolset contains no write/edit/shell tool and cannot read outside the snapshot dir. Findings captured via a typed record_finding tool appending to a slice the stage owns, surviving a Subagent.Run error when findings were already captured."}}
|
||||
{"id":"int-b0d4496e53ae09d1926495a2215fff13","kind":"field_change","created_at":"2026-08-31T11:51:43.090585774Z","actor":"Sam Catlow","issue_id":"archie-core-h019.4","extra":{"field":"status","new_value":"closed","old_value":"open","reason":"StageReview wired into the implement workflow between StageDiffCap and StageOpenPR. Fail-closed: not-run or a confirmed error-level finding parks (store.StatusParked) with findings in Detail; warn/plausible findings never block. Inert unless Repo.ReviewEnabled (new config field, default false); Models[\"reviewer\"] falls back to Models[\"builder\"]. task check clean, golangci-lint clean. Committed 8db8a17."}}
|
||||
{"id":"int-bdb049c55a0a716a843b41f25c644bd7","kind":"field_change","created_at":"2026-09-01T15:18:56.663403225Z","actor":"Sam Catlow","issue_id":"archie-core-atv2","extra":{"field":"status","new_value":"closed","old_value":"open","reason":"Fixed: stopRendering now marks the reply finalized, and appendFallbackLine sends a standalone follow-up message instead of writing into a dead render buffer. Covered by TestLiveReplyMediaFallbackAfterFinalizeSendsFollowUp."}}
|
||||
{"id":"int-79cc9ce98fc14750b24a9590b641f68c","kind":"field_change","created_at":"2026-09-01T15:30:08.83790169Z","actor":"Sam Catlow","issue_id":"archie-core-1786637500752-415-434e5bb9","extra":{"field":"status","new_value":"closed","old_value":"open","reason":"Closed"}}
|
||||
+12
-11
@@ -89,6 +89,7 @@
|
||||
{"_type":"issue","id":"archie-core-g7tm","title":"telegram: finalize the reply through a live context, matching the error branch","description":"The success path calls live.finalize(turnCtx, reply) with the raw cancellable turn context (telegram.go:582), while the error branch two lines above deliberately uses context.WithoutCancel(turnCtx). If /stop lands during finalize, the rich edit, the plain-text retry, and the split-send fallback all abort with context.Canceled (go-telegram/bot builds requests with NewRequestWithContext), so the authoritative reply is never delivered and the live message keeps the ▌ cursor forever. The reply is persisted to the ledger, so the loss is delivery-level — but the user sees a frozen half-answer.\n\nFailure: a long tool-using turn is /stop-ped as it completes; turnCtx is cancelled mid-finalize; every send fails on the dead context; the chat shows the last live frame with a blinking cursor and no final answer.\n\nFound by code-review of PR #476 (feat/467, merged 98e8dda).","design":"live.finalize(context.WithoutCancel(turnCtx), reply) — the one-liner the error branch already uses. Consider making finalize take a live context internally rather than trusting each caller to remember.","acceptance_criteria":"A /stop landing mid-finalize still delivers the authoritative reply and drops the cursor; a test covers the cancellable-context finalize path.","status":"closed","priority":1,"issue_type":"bug","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-14T14:43:08Z","created_by":"Sam Catlow","updated_at":"2026-08-14T15:07:36Z","started_at":"2026-08-14T14:53:28Z","closed_at":"2026-08-14T15:03:28Z","close_reason":"Fixed in internal/channels/telegram/live.go and telegram.go: finalize/abandon strip context cancellation before sending (mirroring each other), ToolCall escapes Markdown metacharacters via bot.EscapeMarkdown, splitLongMessage splits on rune boundaries. All new/changed tests verified red-then-green; task check clean; adversarial review pass found no blocking issues.","external_ref":"https://github.com/samcharles93/archie-core/issues/484","labels":["code-review","streaming","telegram"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-tb3a","title":"Fix streaming review blockers for inline tool calls","description":"Resolve the four evidence-gated findings in merged inline tool-call streaming: preserve the authoritative Telegram reply when final edits fail; keep Telegram network I/O off the model FullStream consumer and cancellable; correlate tool inputs with results and render a bounded redacted parameter summary; snapshot show_tool_calls per live reply to avoid reload races. Add regression tests for every path and run task check plus lint/race validation.","acceptance_criteria":"Regression tests fail on the merged implementation and pass after the fix. Final edit failure falls back to normal split delivery. Delta and ToolCall callbacks do not wait on Telegram HTTP and stop cancels in-flight rendering. Completed tool events include bounded redacted parameters correlated by tool-call ID. Reload cannot race an in-flight reply's show_tool_calls setting. task check passes, golangci-lint reports no new findings, and focused race tests pass.","notes":"Implemented in worktree /tmp/archie-core-tb3a on branch fix/archie-core-tb3a. Added cancellable async Telegram rendering, final edit failure fallback through normal split delivery, call-ID-correlated bounded conservative parameter summaries, atomic per-reply show_tool_calls snapshots, SSE/UI parameter rendering, rebuilt ui/dist, and regression tests. Validation: GOFLAGS=-buildvcs=false task check passed; golangci-lint --new-from-rev=HEAD ./... reported 0 issues; focused race suite passed; fresh adversarial reviewer reported PASS with zero findings. Plain task check is blocked in linked worktrees by Go VCS stamping exit 128, hence buildvcs override.\n2026-08-24 reconciliation: the close reason's 'worktree left uncommitted' is stale. The work did land on main via ba33527 (Merge pull request #477 from samcharles93/fix/archie-core-tb3a); verified on main: gateway/stream.go sensitiveParameterKey+redactToolParameters, telegram/live.go per-reply showToolCalls snapshot, webui/api_chat.go gating. The /tmp/archie-core-tb3a worktree and its branch are gone, as expected.\n","status":"closed","priority":1,"issue_type":"bug","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-13T18:05:35Z","created_by":"Sam Catlow","updated_at":"2026-08-24T07:29:09Z","started_at":"2026-08-13T18:05:39Z","closed_at":"2026-08-13T18:35:26Z","close_reason":"Completed and independently reviewed; worktree left uncommitted per conservative repository policy.","external_ref":"https://github.com/samcharles93/archie-core/issues/485","dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-dzx","title":"Remove duplicate archie-agent safe.directory setup","description":"## Problem\\ncmd/archie-agent/main.go configures git safe.directory inline and then calls markWorktreeSafe, so startup writes the same global git configuration twice.\\n\\n## Scope\\nKeep markWorktreeSafe as the single tested production path and remove the older inline exec.Command block/import. Add regression coverage proving startup wiring has exactly one authoritative safe-directory call.\\n\\n## Acceptance\\n- archie-agent startup invokes markWorktreeSafe exactly once\\n- no second inline safe.directory command remains in main.go\\n- cmd/archie-agent tests and repository quality gate pass","acceptance_criteria":"Single tested safe.directory setup path remains; duplicate inline command is removed; tests and task check pass","notes":"Red proof: TestMainConfiguresWorktreeSafetyOnce failed with 2 setup paths before implementation. Green proof: go test/vet/build ./cmd/archie-agent/... passed; golangci-lint --new-from-rev=HEAD reported 0 issues. Fresh adversarial review found no blockers. Full task check was not run because the working tree contains unrelated in-progress changes and task check rewrites files; scoped package gate used per repository protocol.","status":"closed","priority":1,"issue_type":"bug","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-13T16:10:09Z","created_by":"Sam Catlow","updated_at":"2026-08-13T16:13:48Z","started_at":"2026-08-13T16:10:24Z","closed_at":"2026-08-13T16:13:48Z","close_reason":"Removed duplicate inline safe.directory setup; markWorktreeSafe is now the single tested production path","external_ref":"https://github.com/samcharles93/archie-core/issues/470","dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499161-356-e424e40d.1","title":"setupMemory never calls RegisterExternal; cfg.Memory.Provider is silently inert","description":"internal/app/archied/bootstrap.go:setupMemory's own doc comment says\n\"External providers from config are added via RegisterExternal when\ncfg.Memory.Provider is set\" -- no such branch exists in the function body.\nToday, an operator who sets [memory] provider = \"...\" in config.toml gets\nno error and no effect: the builtin file-backed provider is the only one\nthat ever runs. config.example.toml already documents this as a known gap\n(\"[memory] exists to register an EXTERNAL provider alongside it, but as of\nthis writing that wiring ... is not called\").\n\nSeparately, internal/config.MemoryConfig (config.go:287) has no `toml`\nstruct tag on any field (yaml/json only), so even fixing the wiring call\nmay not bind values from config.toml, the format operators actually use --\nverify this before or as part of the fix.\n\nThis sits ahead of the typed engine-family work (archie-core-...-11-e245c170)\nin the sense that it's a bug in the CURRENT ad-hoc MemoryProvider wiring,\nnot a gap in the future engine contract -- decide whether it's worth fixing\nin the legacy path or superseded once the conforming memory engine family\nlands, rather than fixing both.","status":"open","priority":2,"issue_type":"bug","owner":"sam@catlow.tech","created_at":"2026-09-01T15:50:52Z","created_by":"Sam Catlow","updated_at":"2026-09-01T15:51:00Z","external_ref":"https://github.com/samcharles93/archie-core/issues/649","labels":["epic"],"dependencies":[{"issue_id":"archie-core-1786637499161-356-e424e40d.1","depends_on_id":"archie-core-1786637499161-356-e424e40d","type":"parent-child","created_at":"2026-09-01T15:50:52Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-r3gh","title":"Harden task origin response assertions","description":"Replace brittle prose substring assertions in internal/webui/api_tasks_origin_test.go with stable structural checks for HTTP error responses while preserving origin and media-type coverage.","acceptance_criteria":"The origin test does not assert human-readable response text; rejected cases verify stable HTTP error response structure; focused tests and task check pass.","status":"in_progress","priority":2,"issue_type":"task","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-31T12:49:20Z","created_by":"Sam Catlow","updated_at":"2026-08-31T12:49:27Z","started_at":"2026-08-31T12:49:27Z","lease_expires_at":"2026-08-31T12:54:27Z","heartbeat_at":"2026-08-31T12:49:27Z","dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-h3w2","title":"Migrate email/webhook channels onto domain/messaging types","description":"Update remaining gateway-based channels (email, webhook) to use domain/messaging types, completing the migration described in migration-decisions.md section 2.","status":"open","priority":2,"issue_type":"task","owner":"sam@catlow.tech","created_at":"2026-08-31T10:12:16Z","created_by":"Sam Catlow","updated_at":"2026-08-31T10:12:16Z","labels":["messaging-migration"],"dependencies":[{"issue_id":"archie-core-h3w2","depends_on_id":"archie-core-d9dv","type":"blocks","created_at":"2026-08-31T10:12:24Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-o4e4","title":"Migrate web UI / dashboard chat onto domain/messaging types","description":"Update internal/webui/ chat handling (api_chat.go and friends) to use domain/messaging types once gateway session storage is migrated. Verify SSE streaming and chat history are unaffected.","status":"open","priority":2,"issue_type":"task","owner":"sam@catlow.tech","created_at":"2026-08-31T10:12:15Z","created_by":"Sam Catlow","updated_at":"2026-08-31T10:12:15Z","labels":["messaging-migration"],"dependencies":[{"issue_id":"archie-core-o4e4","depends_on_id":"archie-core-d9dv","type":"blocks","created_at":"2026-08-31T10:12:23Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
|
||||
@@ -192,7 +193,7 @@
|
||||
{"_type":"issue","id":"archie-core-1786637500839-418-cad065a6","title":"SOUL loader with per-turn reload and default fallback","description":"Load the SOUL file from the config dir into the identity slot of the rendered prompt, re-reading it per turn so an edit takes effect on the next message with no restart.\n\n### Acceptance Criteria\n\nSOUL content is rendered into the identity slot ahead of the invariant sections. An absent, empty or unreadable SOUL falls back to the shipped default identity rather than rendering nothing. Editing the file changes the next turn's prompt with no restart. A read error is logged once and degrades to the default; it never fails the turn. Size is bounded so an oversized SOUL cannot crowd out the rules.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:59:24Z","updated_at":"2026-08-08T14:59:24Z","external_ref":"https://github.com/samcharles93/archie-core/issues/439","source_system":"github:https://github.com/samcharles93/archie-core/issues/439:439","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500805-417-67db48ed","title":"Regression tests: Container pool — volume lifecycle, error paths, concurrent pool access","description":"internal/container/pool_test.go has 3 tests (grace period, JSON output, volume file). Missing: volume cleanup on error, concurrent pool access, container timeout during task, invalid task payload handling, Docker daemon unavailable error path.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:19Z","updated_at":"2026-08-08T14:59:22Z","closed_at":"2026-08-08T14:59:23Z","external_ref":"https://github.com/samcharles93/archie-core/issues/438","source_system":"github:https://github.com/samcharles93/archie-core/issues/438:438","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500776-416-600a66e5","title":"Surprisal sampling: embedding-backed strategies","description":"Implement the embedding-backed surprisal strategy behind the Sampler seam that wave 1 of the curator epic ships (cheap strategies: recency, random, all, staleness proxy). Surprisal scores memories by how surprising they are so the sampled reasoner's agentic budget is spent on the memories most likely to matter. DEFERRED by design: gated behind the embeddings capability (sibling issue 'Embeddings capability (prerequisite for surprisal sampling)') — pick up only after that lands.\n\n### Acceptance Criteria\n\nA Sampler implementation behind the same interface as the cheap strategies; selection is deterministic given fixed inputs. Algorithm: sample up to a candidate cap; build a k-NN graph over embedding vectors; row-normalize the adjacency with self-loops; power-iterate to the stationary distribution; surprisal = -log(stationary probability of the nearest point); select the highest-surprisal items. Degradation: content with no embedding is skipped; an embedding failure aborts sampling and falls back to free exploration - a sampling failure must never fail or crash a curator pass. Cost: the graph build is quadratic in the candidate cap; the cap is a named constant with a documented bound. Tests use synthetic vectors via httptest, never a live embedding service.","status":"open","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:16Z","updated_at":"2026-08-08T14:59:16Z","external_ref":"https://github.com/samcharles93/archie-core/issues/437","source_system":"github:https://github.com/samcharles93/archie-core/issues/437:437","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500752-415-434e5bb9","title":"Embeddings capability (prerequisite for surprisal sampling)","description":"Surprisal-based sampling (curator epic DESIGN) requires computing vector embeddings of memory content so memories can be scored by how surprising they are. Archie has no embedding capability today; the epic gates any surprisal-style sampling behind adding one. This issue tracks that prerequisite so it can be picked up independently of the curator epic's main work. The consumer is the embedding-backed surprisal sampler (sibling issue 'Surprisal sampling: embedding-backed strategies'). Embedding consumers must degrade gracefully: content with no embedding is skipped, and an embedding failure aborts sampling in favour of free exploration rather than crashing the pass.\n\n### Acceptance Criteria\n\nAn embedding client exists behind a narrow typed interface usable by in-process daemon components (curators first, other consumers later). Provider/model is config-driven (providers table + a role in [models]). A missing credential degrades per the credential-missing-degrades-not-fatal rule: capability reports unavailable, daemon still starts. Network calls carry timeouts; failures surface as reported errors, never a crash or a failed chat turn. Tests use httptest and never require a live embedding service.","status":"open","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:13Z","updated_at":"2026-08-08T14:59:13Z","external_ref":"https://github.com/samcharles93/archie-core/issues/436","source_system":"github:https://github.com/samcharles93/archie-core/issues/436:436","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500752-415-434e5bb9","title":"Embeddings capability (prerequisite for surprisal sampling)","description":"Surprisal-based sampling (curator epic DESIGN) requires computing vector embeddings of memory content so memories can be scored by how surprising they are. Archie has no embedding capability today; the epic gates any surprisal-style sampling behind adding one. This issue tracks that prerequisite so it can be picked up independently of the curator epic's main work. The consumer is the embedding-backed surprisal sampler (sibling issue 'Surprisal sampling: embedding-backed strategies'). Embedding consumers must degrade gracefully: content with no embedding is skipped, and an embedding failure aborts sampling in favour of free exploration rather than crashing the pass.\n\n### Acceptance Criteria\n\nAn embedding client exists behind a narrow typed interface usable by in-process daemon components (curators first, other consumers later). Provider/model is config-driven (providers table + a role in [models]). A missing credential degrades per the credential-missing-degrades-not-fatal rule: capability reports unavailable, daemon still starts. Network calls carry timeouts; failures surface as reported errors, never a crash or a failed chat turn. Tests use httptest and never require a live embedding service.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:13Z","updated_at":"2026-09-01T15:30:09Z","closed_at":"2026-09-01T15:30:09Z","close_reason":"Closed","external_ref":"https://github.com/samcharles93/archie-core/issues/436","source_system":"github:https://github.com/samcharles93/archie-core/issues/436:436","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500725-414-9d66a52b","title":"[EPIC] Curator engine: persistent memory-reasoning agent loops","description":"Introduce curators as their own engine family: long-running agent loops that reason over memory and perform small maintenance tasks, rather than one hardcoded background job. A curator is defined by a check-in interval, a tool set, and (once the memory engine family lands) an attached memory engine. It runs persistently while there is input and backs off to a cooldown when idle. Ship two reference curators -- a skill curator and a persistent-session-memory curator -- so the interface is proven against two consumers rather than designed around one.\n\n### Design\n\nThe design splits a cheap hot-path extractor from an expensive sampled reasoner, so agentic cost is never paid per message. Sampling chooses which memories deserve attention via surprisal-based scoring, implemented over embeddings through interchangeable strategies including a k-NN graph random walk. Archie has no embedding capability today, so any surprisal-style sampling is gated behind adding one.\n\n### Acceptance Criteria\n\nCurator is a conforming engine family per ARCHITECTURE.md 'Plugin engine rule (strict)': typed contract, owning registry/manager, start/health/stop, narrow typed host access. Curators are registered and configured declaratively, not hardcoded. The runtime supports persistent operation under load and cooldown when idle, with intervals configurable per curator. Both reference curators run against the shipped defaults. Curator activity is observable: what ran, when, what it changed, and why. A curator failure is isolated and never takes down the daemon or a chat turn.","status":"open","priority":2,"issue_type":"epic","created_at":"2026-08-08T14:59:10Z","updated_at":"2026-08-08T15:34:16Z","external_ref":"https://github.com/samcharles93/archie-core/issues/435","source_system":"github:https://github.com/samcharles93/archie-core/issues/435:435","metadata":null,"labels":["epic"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500699-413-301bc587","title":"Regression tests: Yaegi extract packages smoke tests (wfextract, gateextract, pluginextract)","description":"Four extract packages have zero tests: internal/workflow/wfextract, internal/gate/gateextract, internal/plugin/pluginextract, and internal/workflow/wfextract. These are Yaegi symbol extractors that generate Go source exposing package symbols to Yaegi scripts. They fail loudly at build time if broken, but minimal smoke tests would catch regressions in the export format (correct symbol names, expected types, no missing exports).","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:05Z","updated_at":"2026-08-08T14:59:07Z","closed_at":"2026-08-08T14:59:08Z","external_ref":"https://github.com/samcharles93/archie-core/issues/434","source_system":"github:https://github.com/samcharles93/archie-core/issues/434:434","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637500667-412-04faa3e7","title":"generalize Provider.APIKeyEnv / NATS.TokenEnv / Telegram.TokenEnv to SecretRef","description":"Follow-up to archie-core-xec, which only migrated Forge.TokenEnv (self-contained call site). Provider.APIKeyEnv is forwarded THREE different ways that all assume the secret is reachable by name in the daemon's own env, not resolved centrally:\n1. agentexec.NewRuntime -\u003e passes the env-var NAME into ai-sdk/runtime's AuthConfig.APIKeyEnv, which lazily os.Getenv's it later (external module, cannot change its contract).\n2. agentexec.WorkerEnvironment / subprocess.go WorkerEnvironment() -\u003e allowlist-forwards the NAME into a subprocess's environment for archie-agent to read itself.\n3. daemon.containerEnv() -\u003e already eagerly resolves via os.Getenv and injects NAME=value into a container env (this one generalizes cleanly).\n\nMigrating Provider.APIKeyEnv to SecretRef requires switching cases 1 and 2 from name-forwarding to eager-resolve-via-secret.Registry-then-inject-value (ai-sdk/runtime.AuthConfig.APIKey literal field already supports this per its ResolveAPIKey fallback chain -- no ai-sdk/runtime changes needed, just stop relying on its lazy os.Getenv path for non-env engines). Also migrate NATSConfig.TokenEnv and TelegramConfig.TokenEnv (each has a single simple os.Getenv call site, lower risk than Provider).\n\n### Design\n\nsecret.Registry must be threaded into agentexec.NewRuntime, WorkerEnvironment, and daemon.containerEnv (or a shared secret-resolution helper) so eager-resolve-then-inject replaces name-forwarding for engines other than the built-in env engine.\n\n### Acceptance Criteria\n\nProvider.APIKeyEnv, NATSConfig.TokenEnv, TelegramConfig.TokenEnv replaced with SecretRef; all three consumption paths (in-process runtime, subprocess worker env, container env) resolve via secret.Registry; task check passes; adversarial review finds zero surviving findings; explicit test proving a non-env engine (e.g. a fake test engine) correctly delivers a provider API key to agentexec.NewRuntime via the AuthConfig.APIKey literal field, not APIKeyEnv.","status":"open","priority":2,"issue_type":"task","created_at":"2026-08-08T14:59:02Z","updated_at":"2026-08-08T15:35:20Z","external_ref":"https://github.com/samcharles93/archie-core/issues/433","source_system":"github:https://github.com/samcharles93/archie-core/issues/433:433","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -221,7 +222,7 @@
|
||||
{"_type":"issue","id":"archie-core-1786637500022-389-a765fd7a","title":"Telegram: component-aware /update with approve or defer","description":"Add an /update control flow that checks for available releases, presents separate archied gateway and archie-agent runtime versions/changelog sections, and lets an authorized user approve or defer. Keep discovery and execution behind typed interfaces so Docker/Watchtower, package-manager, and other installations can supply their own updater without the Telegram gateway assuming a deployment mechanism.\n\n### Acceptance Criteria\n\n/update reports whether gateway and runtime updates are available, with separate version and changelog sections and explicit no-change wording. Telegram provides Approve and Defer actions restricted to authorized users. Deferral is persisted and suppresses repeated prompts according to an explicit policy. Approval calls a typed installation updater and reports progress/failure without claiming success early. No Watchtower-only or Docker-only behavior exists in the gateway package. Commands, menu, help, callback authorization, concurrency, persistence, and failure paths are tested.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:57:17Z","updated_at":"2026-08-08T14:57:20Z","closed_at":"2026-08-08T14:57:21Z","external_ref":"https://github.com/samcharles93/archie-core/issues/410","source_system":"github:https://github.com/samcharles93/archie-core/issues/410:410","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499992-388-15ec94cd","title":"fix(memory): startup accepted an unusable memory directory","description":"cmd/archied/main.go builds the builtin memory provider at filepath.Join(cfg.WorkDir, \"memory\") and logs 'memory manager started' regardless of whether that path exists or is writable. builtin.NewStore treats ENOENT as an empty document and Reload does the same, so a missing directory is indistinguishable from a memory with nothing in it yet.\n\nFound live: carina's config was copied to the laptop with work_dir = /var/lib/archie/work, a container path that does not exist here. archied logged 'memory manager started dir=/var/lib/archie/work/memory' and ran with no memory at all. The failure would only have surfaced at the agent's first memory write, when writeFileAtomic's MkdirAll hit EACCES on /var/lib -- mid-conversation, as a tool error, after startup said it was fine.\n\nFixed: builtin.New now does MkdirAll(cfg.Dir) and returns an error the operator can act on, so an unusable path fails at startup where the existing error path already exits non-zero. New's doc claimed 'the provider is ready for use as soon as New returns'; it now is.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:57:12Z","updated_at":"2026-08-08T14:57:14Z","closed_at":"2026-08-08T14:57:15Z","external_ref":"https://github.com/samcharles93/archie-core/issues/409","source_system":"github:https://github.com/samcharles93/archie-core/issues/409:409","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499965-387-1f257fda","title":"Audit archie-core repo for bugs/improvements, file as Gitea issues","description":"Fan out exploration subagents across internal/ subsystems to find bugs, tech debt, and enhancement opportunities. Each confirmed finding gets filed as a Gitea issue on git.catlow.cloud/sam/archie-core.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:57:06Z","updated_at":"2026-08-08T14:57:09Z","closed_at":"2026-08-08T14:57:10Z","external_ref":"https://github.com/samcharles93/archie-core/issues/408","source_system":"github:https://github.com/samcharles93/archie-core/issues/408:408","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499933-386-83646cce","title":"Curator contract, registry, and internal/curator package","description":"Stand up internal/curator as a package in its own right: the typed curator contract, the owning registry that controls registration and discovery, and the definition of a curator's declared shape (check-in interval, tool set, attached memory engine).\n\n### Acceptance Criteria\n\ninternal/curator owns a typed contract exposing real curator operations plus start/health/stop. A registry controls registration, discovery, and shutdown ordering, and defines failure isolation. Curators receive a narrow typed registrar and only the host services their declared capability requires -- never *daemon.Daemon or an untyped hook map. internal/plugin/architecture_test.go recognises curator as a conforming engine family.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:57:04Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator (contract.go, registry.go, registrar.go) already satisfies every acceptance criterion. CuratorEngine is a typed contract with Name/Version/Manifest/Check/Pass/Bind plus embedded Lifecycle (start/health/stop). Registry (registry.go) controls Register/Start/Health/Stop with dependency-free single ownership, panic-isolated safeCheck/safePass/bindSafely. Registrar (registrar.go) is narrow and fully typed (LLMRunner, ToolBuilder, SkillStore, EventSink, Clock) -- no *daemon.Daemon, no untyped hook map, confirmed by grep. go test ./internal/plugin/... -run TestEveryEngineInterfaceHasTypedCapabilityAndFamilyOwner -v shows domain/curator.CuratorEngine PASS as a recognised conforming engine family. go test ./internal/domain/curator/... -race -count=1 is fully green. Note: package lives at internal/domain/curator (per the domain migration), not literally internal/curator as the title says -- that's the current architecture, not a gap.","external_ref":"https://github.com/samcharles93/archie-core/issues/407","source_system":"github:https://github.com/samcharles93/archie-core/issues/407:407","metadata":null,"labels":["enhancement","in-progress"],"dependencies":[{"issue_id":"archie-core-1786637499933-386-83646cce","depends_on_id":"archie-core-1786637490069-11-e245c170","type":"related","created_at":"2026-08-15T12:49:33Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499933-386-83646cce","title":"Curator contract, registry, and internal/curator package","description":"Stand up internal/curator as a package in its own right: the typed curator contract, the owning registry that controls registration and discovery, and the definition of a curator's declared shape (check-in interval, tool set, attached memory engine).\n\n### Acceptance Criteria\n\ninternal/curator owns a typed contract exposing real curator operations plus start/health/stop. A registry controls registration, discovery, and shutdown ordering, and defines failure isolation. Curators receive a narrow typed registrar and only the host services their declared capability requires -- never *daemon.Daemon or an untyped hook map. internal/plugin/architecture_test.go recognises curator as a conforming engine family.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:57:04Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator (contract.go, registry.go, registrar.go) already satisfies every acceptance criterion. CuratorEngine is a typed contract with Name/Version/Manifest/Check/Pass/Bind plus embedded Lifecycle (start/health/stop). Registry (registry.go) controls Register/Start/Health/Stop with dependency-free single ownership, panic-isolated safeCheck/safePass/bindSafely. Registrar (registrar.go) is narrow and fully typed (LLMRunner, ToolBuilder, SkillStore, EventSink, Clock) -- no *daemon.Daemon, no untyped hook map, confirmed by grep. go test ./internal/plugin/... -run TestEveryEngineInterfaceHasTypedCapabilityAndFamilyOwner -v shows domain/curator.CuratorEngine PASS as a recognised conforming engine family. go test ./internal/domain/curator/... -race -count=1 is fully green. Note: package lives at internal/domain/curator (per the domain migration), not literally internal/curator as the title says -- that's the current architecture, not a gap.","external_ref":"https://github.com/samcharles93/archie-core/issues/407","source_system":"github:https://github.com/samcharles93/archie-core/issues/407:407","metadata":null,"labels":["enhancement","in-progress"],"dependencies":[{"issue_id":"archie-core-1786637499933-386-83646cce","depends_on_id":"archie-core-1786637490069-11-e245c170","type":"related","created_at":"2026-08-15T12:49:33Z","created_by":"Sam Catlow","metadata":"{}"},{"issue_id":"archie-core-1786637499933-386-83646cce","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:26Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499903-385-ed6b4aa1","title":"Telegram: remove task approval from the public command surface","description":"Existing /approve [identity=name] \u003ctask-id\u003e was inherited as chat-task lifecycle control, but Sam explicitly states task approvals are not a product requirement. Reserve /approve exclusively for dangerous-command approval. Remove task approval from Telegram command menu/help/router/public UX; retain or remove internal waiting_human task mechanics only after verifying workflow callers. Do not repurpose the command until the dangerous-command approval model (archie-core-alm.6) exists.\n\n### Acceptance Criteria\n\nTelegram no longer advertises or accepts task approval via /approve; no documentation/help describes task approvals; task lifecycle code is not broken for existing non-Telegram workflow paths; /approve remains unclaimed or is implemented only for the approved dangerous-command design.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:56:58Z","updated_at":"2026-08-08T14:57:01Z","closed_at":"2026-08-08T14:57:02Z","external_ref":"https://github.com/samcharles93/archie-core/issues/406","source_system":"github:https://github.com/samcharles93/archie-core/issues/406:406","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499872-384-7952eb47","title":"chat: mid-conversation context loss, undiagnosed","description":"Reported from a real Telegram transcript (29 Jul, ~09:47 AEST / 23:47 UTC 28\nJul). Turn 3 of a coherent conversation was answered with a fresh greeting as\nthough no history existed; turn 4 was coherent again.\n\nRuled out by evidence:\n- Not two processes racing the bot token. docker compose ps on carina shows a\n single archied container.\n- Not a restart. That container had 43h uptime spanning the whole exchange.\n- Not history compression. DefaultCompressionConfig protects the first 3 and\n last 20 messages and triggers at 50% of a 128k window; there were three\n messages.\n- Not a canned greeting. No such string exists in the codebase, so the reply\n was model-generated from little or no history.\n- Not the session-key namespace split fixed in 96ab167: the deployed image\n predates the session lifecycle commands entirely.\n\nBlocked on evidence, now unblocked: chat turns were completely unlogged (76\nlog lines covered two days, none a conversation). 96ab167 adds a \"chat turn\"\nlog line carrying session id, channel, thread and history depth.\n\nNext step when it recurs: check the chat turn log for that message. A session\nid differing from neighbouring turns points at key resolution (e.g. ThreadID\nappearing on one message); history_messages of 0 or 1 with a stable session id\npoints at the store read path.\n\nDo not close this by guessing. It needs one reproduction with the new log line.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:56:55Z","updated_at":"2026-08-17T14:34:48Z","closed_at":"2026-08-17T14:34:48Z","close_reason":"Confirmed fixed by Sam; no longer reproducible","external_ref":"https://github.com/samcharles93/archie-core/issues/405","source_system":"github:https://github.com/samcharles93/archie-core/issues/405:405","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499843-383-4eb46948","title":"Flaky: TestRunTaskExecutesBootstrapWorkflowEndToEnd races on NATS responders","description":"cmd/archie-agent. Intermittently fails with ParkReason 'stage prepare: archie.worktree.prepare: nats: no responders available for request'; passes on re-run (-count=3 clean). The test does not wait for the worktreerpc subscription to be established before dispatching. A flaky test in task check is poison: it trains everyone to re-run instead of read, which is how a real regression gets waved through.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:56:52Z","updated_at":"2026-08-15T12:50:59Z","closed_at":"2026-08-15T12:50:59Z","close_reason":"Fixed: internal/natsrpc/natsrpc.go:118-121 RegisterAll already calls nc.Flush() before returning, exactly matching the fix CLAUDE.md documents ('natsrpc.RegisterAll now flushes before returning -- this was the cause of the long-standing TestRunTaskExecutesBootstrapWorkflowEndToEnd CI flake'). Verified: go test ./internal/natsrpc/... ./internal/app/agentworker/... ./internal/daemon/... -race -count=3 is clean, no flakes.","external_ref":"https://github.com/samcharles93/archie-core/issues/404","source_system":"github:https://github.com/samcharles93/archie-core/issues/404:404","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -231,7 +232,7 @@
|
||||
{"_type":"issue","id":"archie-core-1786637499722-379-670f62fc","title":"Phase 2: mechanical per-package fixes (~40 issues)","description":"forcetypeassert (24, internal/tools 20 + internal/memory 2, nearly all tests), predeclared (9, webui/ratelimit/memory), leftover unconvert. Purely mechanical, package-local, no design judgement. THE ONLY PHASE WORTH PARALLELISING. Partition by package: tools / webui / memory+ratelimit / secret. Use git worktree isolation -- other sessions edit this repo concurrently and stale diagnostics are a known hazard. Each worker verifies scoped: go test ./internal/X/... plus golangci-lint run ./internal/X/...","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:56:34Z","updated_at":"2026-08-08T14:56:36Z","closed_at":"2026-08-08T14:56:37Z","external_ref":"https://github.com/samcharles93/archie-core/issues/400","source_system":"github:https://github.com/samcharles93/archie-core/issues/400:400","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499692-378-45486e7d","title":"Phase 1: autofix pass (~35 issues)","description":"golangci-lint run --fix plus the formatters. Covers errorlint, fatcontext, gocritic, sloglint, usestdlibvars, usetesting, gci, gofumpt, goimports. Zero judgement. MUST run first: it touches formatting across every file and would conflict with anything in flight. Single command, single commit, verify with task check.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:56:28Z","updated_at":"2026-08-08T14:56:31Z","closed_at":"2026-08-08T14:56:32Z","external_ref":"https://github.com/samcharles93/archie-core/issues/399","source_system":"github:https://github.com/samcharles93/archie-core/issues/399:399","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499664-377-c2818afd","title":"[EPIC] Clear the 160 golangci-lint findings","description":"Config landed in 9298afd; the gate is RED until this epic closes. Four phases, ordered by risk and by what invalidates what. Structural refactor MUST be last: it moves line numbers and invalidates everything in flight.","status":"open","priority":2,"issue_type":"epic","created_at":"2026-08-08T14:56:25Z","updated_at":"2026-08-08T15:34:26Z","external_ref":"https://github.com/samcharles93/archie-core/issues/398","source_system":"github:https://github.com/samcharles93/archie-core/issues/398:398","metadata":null,"labels":["epic"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499636-376-2f94082e","title":"Bind curators to memory engines","description":"Wire the attached-memory-engine part of a curator's declared shape once the memory engine family exists, so a curator reads and writes through the engine contract rather than a concrete store.\n\n### Acceptance Criteria\n\nA curator declares which memory engine it uses and receives it as a typed dependency. Swapping the configured engine changes where a curator's memories land with no change to the curator. A curator with no declared memory engine is a valid configuration and is rejected only if it attempts a memory operation.","notes":"Blocked on its own stated prerequisite, which is also still open: archie-core-1786637490069-11-e245c170 (Memory engine contract and owning registry). Current state: Manifest.MemoryEngine (internal/domain/curator/contract.go:59-63) is only a declared string name; Registrar (registrar.go) has no memory-engine field at all, so nothing actually binds an engine instance to a curator yet -- registry.go:128's agentic() check only tests MemoryEngine != \"\", it doesn't inject anything. Do not start this until the memory engine contract exists; there is nothing to bind to yet.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:56:23Z","updated_at":"2026-08-15T12:55:30Z","external_ref":"https://github.com/samcharles93/archie-core/issues/397","source_system":"github:https://github.com/samcharles93/archie-core/issues/397:397","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637499636-376-2f94082e","depends_on_id":"archie-core-1786637490069-11-e245c170","type":"blocks","created_at":"2026-08-15T12:55:30Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499636-376-2f94082e","title":"Bind curators to memory engines","description":"Wire the attached-memory-engine part of a curator's declared shape once the memory engine family exists, so a curator reads and writes through the engine contract rather than a concrete store.\n\n### Acceptance Criteria\n\nA curator declares which memory engine it uses and receives it as a typed dependency. Swapping the configured engine changes where a curator's memories land with no change to the curator. A curator with no declared memory engine is a valid configuration and is rejected only if it attempts a memory operation.","notes":"Blocked on its own stated prerequisite, which is also still open: archie-core-1786637490069-11-e245c170 (Memory engine contract and owning registry). Current state: Manifest.MemoryEngine (internal/domain/curator/contract.go:59-63) is only a declared string name; Registrar (registrar.go) has no memory-engine field at all, so nothing actually binds an engine instance to a curator yet -- registry.go:128's agentic() check only tests MemoryEngine != \"\", it doesn't inject anything. Do not start this until the memory engine contract exists; there is nothing to bind to yet.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:56:23Z","updated_at":"2026-08-15T12:55:30Z","external_ref":"https://github.com/samcharles93/archie-core/issues/397","source_system":"github:https://github.com/samcharles93/archie-core/issues/397:397","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637499636-376-2f94082e","depends_on_id":"archie-core-1786637490069-11-e245c170","type":"blocks","created_at":"2026-08-15T12:55:30Z","created_by":"Sam Catlow","metadata":"{}"},{"issue_id":"archie-core-1786637499636-376-2f94082e","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:27Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499610-375-5e43fe37","title":"Consolidate duplicated NATS RPC client/server boilerplate","description":"internal/storerpc/storerpc.go, internal/worktreerpc/worktreerpc.go, internal/forgerpc/forgerpc.go each reimplement near-identical timeout/marshal/error-envelope helpers for NATS request/reply RPC. Extract shared helpers into a common package.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:56:17Z","updated_at":"2026-08-08T14:56:20Z","closed_at":"2026-08-08T14:56:21Z","external_ref":"https://github.com/samcharles93/archie-core/issues/396","source_system":"github:https://github.com/samcharles93/archie-core/issues/396:396","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499585-374-bf060b8b","title":"infrastructure/eventbus/nats: relocate NATS transport","description":"Create internal/infrastructure/eventbus/nats as a properly structured replacement for internal/nats. This is NOT a mechanical move: the existing code has no interfaces, leaks vendor types (Conn() *nats.Conn, PublishMsg(*nats.Msg), Fetch() jetstream.Msg) into all four consumers, uses a 7-positional-parameter publish with stringly-typed comma-separated labels, stores an unused logger, returns (nil, nil) for 'no message', and hardcodes stream/consumer/timeout config. Rebuild it with: separate files per concern; a Message wrapper so jetstream types stop crossing the boundary; Publisher/Consumer interfaces; a Config struct; sentinel errors; consistent %w wrapping; real logging. Then update cmd/archied, cmd/archie-agent, internal/daemon and internal/agentexec to the new API.\n\n### Acceptance Criteria\n\ninternal/nats no longer exists; internal/infrastructure/eventbus/nats has its contents. All imports updated. task check clean. Container and NATS behaviour unchanged.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:56:12Z","updated_at":"2026-08-08T14:56:15Z","closed_at":"2026-08-08T14:56:16Z","external_ref":"https://github.com/samcharles93/archie-core/issues/395","source_system":"github:https://github.com/samcharles93/archie-core/issues/395:395","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499562-373-11ad6000","title":"fix(nell): counters squat on the SDK-reserved meta: prefix, blocking all writes after restart","description":"TOTAL OUTAGE. No task can be queued and no event recorded; survives restarts.\n\ninternal/nell/adapter.go:946,952 store auto-increment counters at meta:task_id_counter and meta:event_id_counter. The NellDB SDK reserves the meta: prefix for its own bookkeeping (sdk/replicate.go:841 isInternalID = id[:5]==\"meta:\").\n\nDocDB.reindex() skips internal ids when rebuilding the in-memory rev cache (sdk/docdb.go:128-137), but the record stays in the log. So after a restart Get returns the counter doc WITH its _rev while d.revs has no entry, and putIn rejects exactly that: 'incomingRev != \"\" \u0026\u0026 !exists' -\u003e ErrConflict (sdk/docdb.go:213-217). Permanent.\n\nHits nextTaskID (every EnqueueIssue) and nextEventID (every InsertEvent). processNATSTask Naks on error so NATS redelivers forever.\n\nReproduced: internal/nell/counter_repro_test.go - open, insert, close, reopen, insert -\u003e ErrConflict in 7ms with no daemon/NATS/forge.\n\nSECOND DEFECT: isInternalID exists to exclude records from replication, so the two carina instances never converge their counters and would allocate COLLIDING task IDs. Masked only because allocation currently fails outright.\n\nFIX: move the keys out of the reserved namespace (counter:task_id / counter:event_id) AND migrate on open - the old doc is readable (only Put fails), and its next_id is load-bearing because restarting at 1 would reissue live task IDs. Also teach isMetaKey (adapter.go:803) to skip the new prefix so scans do not parse a counter as a task.\n\nREJECTED: stripping _rev before Put is one line and works, but leaves the prefix squat, leaves counters unreplicated, and turns a loud failure into silent last-write-wins.\n\nFull writeup: docs/incident-nelldb-counter-conflict.md","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:56:07Z","updated_at":"2026-08-08T14:56:09Z","closed_at":"2026-08-08T14:56:10Z","external_ref":"https://github.com/samcharles93/archie-core/issues/394","source_system":"github:https://github.com/samcharles93/archie-core/issues/394:394","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -249,11 +250,11 @@
|
||||
{"_type":"issue","id":"archie-core-1786637499268-361-285a7a81","title":"feat(webui): logs section - filterable, live-tailing","description":"There is no way to see archied's logs outside journalctl. Every event already flows through events.Bus into the store and out over SSE, so this is mostly presentation over existing data. Filterable by level and component, live-tailing the SSE stream, with history from the store for anything before the dashboard connected.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:55:09Z","updated_at":"2026-08-08T15:33:52Z","closed_at":"2026-08-08T15:33:53Z","external_ref":"https://github.com/samcharles93/archie-core/issues/382","source_system":"github:https://github.com/samcharles93/archie-core/issues/382:382","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499248-360-a6ac2616","title":"Regression tests: NATS RPC — request timeout, reply routing, error propagation","description":"internal/natsrpc/natsrpc.go has only 1 test (envelope round trip). Missing: request timeout handling, reply-to routing correctness, error propagation through the envelope, concurrent request isolation, reconnection behavior.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:55:03Z","updated_at":"2026-08-08T14:55:06Z","closed_at":"2026-08-08T14:55:07Z","external_ref":"https://github.com/samcharles93/archie-core/issues/381","source_system":"github:https://github.com/samcharles93/archie-core/issues/381:381","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499227-359-c3682a55","title":"Audit the remaining golangci-lint autofix changes in 6d4e0be","description":"Commit 6d4e0be applied a golangci-lint --fix batch across 8 files. One of the eight was a behaviour regression: the errorlint fix in internal/tools/builtin/grep.go rewrote 'if exitErr, ok := err.(*exec.ExitError); ok \u0026\u0026 exitErr.ExitCode() == 1' into 'errors.As(err, \u0026exitErr)' and dropped the exit-code check, so any ripgrep failure (bad pattern, unreadable path) was reported to the model as 'no matches found' -- the agent would state a file contains nothing on the strength of a search that never ran. Fixed in 782510d. The other seven changes were eyeballed as benign (naked return expanded, err != ctx.Err() -\u003e errors.Is) but were never verified individually. A one-in-eight regression rate does not justify trusting the remainder on inspection alone.\n\n### Acceptance Criteria\n\nEach remaining changed hunk in 6d4e0be is reviewed against its pre-fix behaviour, specifically for predicates that were narrowed or widened. internal/channels/telegram/model.go, internal/infrastructure/configuration/discovery.go, internal/tools/builtin/find.go, internal/tools/builtin/shell.go and the touched test files are covered. Any behaviour change found gets a regression test in the style of TestGrepReportsSearchFailureRatherThanNoMatches.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:54:58Z","updated_at":"2026-08-08T15:34:09Z","closed_at":"2026-08-08T15:34:10Z","external_ref":"https://github.com/samcharles93/archie-core/issues/380","source_system":"github:https://github.com/samcharles93/archie-core/issues/380:380","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499206-358-54f29252","title":"Skill curator (reference implementation)","description":"Ship a curator that maintains Archie's skills: reviewing, pruning, and improving skill definitions over time. One of the two consumers that validate the curator contract is general enough.\n\n### Acceptance Criteria\n\nRegisters and runs through the curator registry with no curator-specific special-casing in the runtime. Its changes to skills are observable and attributable to the curator. It operates within its declared tool set and cannot reach tools it did not declare.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:54:55Z","updated_at":"2026-08-08T14:54:55Z","external_ref":"https://github.com/samcharles93/archie-core/issues/379","source_system":"github:https://github.com/samcharles93/archie-core/issues/379:379","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499206-358-54f29252","title":"Skill curator (reference implementation)","description":"Ship a curator that maintains Archie's skills: reviewing, pruning, and improving skill definitions over time. One of the two consumers that validate the curator contract is general enough.\n\n### Acceptance Criteria\n\nRegisters and runs through the curator registry with no curator-specific special-casing in the runtime. Its changes to skills are observable and attributable to the curator. It operates within its declared tool set and cannot reach tools it did not declare.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:54:55Z","updated_at":"2026-08-08T14:54:55Z","external_ref":"https://github.com/samcharles93/archie-core/issues/379","source_system":"github:https://github.com/samcharles93/archie-core/issues/379:379","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637499206-358-54f29252","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:28Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499183-357-e3f7603f","title":"fix(gateway): redelivery dedups the stored turn but still re-runs the model","description":"archie-core-963 made SaveMessage idempotent, but nothing upstream consults the result. cmd/archied/telegram_setup.go:158 respond() calls SaveMessage (now a no-op on redelivery) then unconditionally loads history, calls the model, and saves the reply -- which carries no SourceID, so it always gets a fresh random ID and always appends.\n\nCONFIRMED at store level on both backends: redelivering tg-1 leaves [ping, pong, pong], plus a second billed model turn with any tool side effects executed twice.\n\nThe bead's stated problem ('a re-delivered Telegram update appends a duplicate turn') is half fixed. SaveMessage returns no already-seen signal for the caller to act on.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:54:50Z","updated_at":"2026-08-08T14:54:52Z","closed_at":"2026-08-08T14:54:53Z","external_ref":"https://github.com/samcharles93/archie-core/issues/378","source_system":"github:https://github.com/samcharles93/archie-core/issues/378:378","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499161-356-e424e40d","title":"[EPIC] Pluggable memory engine: typed family + external providers","description":"Make the memory subsystem a pluggable engine family so an external memory backend (Honcho, or a future alternative) can be bolted on without touching the agent loop, while the file-writing store ships as the default. internal/memory already has a Manager and lifecycle hooks (TurnStartHook, SessionEndHook, SessionSwitchHook); this epic promotes that into a conforming engine family per the strict plugin engine rule in ARCHITECTURE.md rather than inventing a parallel abstraction.\n\n### Design\n\nConstraint from the plugin engine rule: 'engine' requires a typed domain contract, an owning Registry or Manager, explicit lifecycle/health, and narrow typed registrars -- providers never receive *daemon.Daemon or untyped hook maps. The contract must express real memory operations (write observation, query, list, forget), not just Name/Version.\n\n### Acceptance Criteria\n\nMemory is a typed engine family with an owning registry/manager, start/health/stop lifecycle, and narrow typed host access, satisfying every invariant in ARCHITECTURE.md 'Plugin engine rule (strict)' and passing internal/plugin/architecture_test.go. The builtin file-writing store is the shipped default engine and remains the behaviour when nothing is configured. At least one external engine is implementable against the contract without changes to the agent loop or gateway. Engine selection and per-engine settings are configurable, validated, and documented. Failure of a non-default engine degrades to a reported error rather than taking down a turn.","status":"open","priority":2,"issue_type":"epic","created_at":"2026-08-08T14:54:47Z","updated_at":"2026-08-08T14:54:47Z","external_ref":"https://github.com/samcharles93/archie-core/issues/377","source_system":"github:https://github.com/samcharles93/archie-core/issues/377:377","metadata":null,"labels":["epic"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499136-355-c766a8ee","title":"Quiet autonomous execution and correct forge linkage","description":"Reduce Archie forge noise and make issue/branch/PR lifecycle authoritative. Automatic plan/progress/retry/state-label comments must be disabled; human decisions must remain available through messaging and the Web UI; pushed task branches must be linked to their source issues through the forge API before PR creation.\n\n### Acceptance Criteria\n\nNo automatic plan/progress/retry/state-label comments are emitted; configured human decisions remain actionable through messaging or Web UI; the source issue has a forge branch link after push; PR creation does not rely on a comment or closing directive for linkage; focused tests cover both in-process and RPC forge paths.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:54:42Z","updated_at":"2026-08-08T14:54:44Z","closed_at":"2026-08-08T14:54:45Z","external_ref":"https://github.com/samcharles93/archie-core/issues/376","source_system":"github:https://github.com/samcharles93/archie-core/issues/376:376","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499114-354-e5b475b4","title":"Persistent session-memory curator (reference implementation)","description":"Ship a curator that maintains durable memory across sessions -- extracting and consolidating what is worth keeping from conversation history into the attached memory engine. The second contract consumer, and the one that exercises the curator/memory-engine binding.\n\n### Acceptance Criteria\n\nRegisters and runs through the registry with no special-casing. Writes through the memory engine contract rather than any concrete store. Extraction runs off the hot path so a chat turn never waits on it. What it wrote and why is inspectable. Honours the trigger-accounting invariant so its own writes do not reschedule it.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:54:39Z","updated_at":"2026-08-08T14:54:39Z","external_ref":"https://github.com/samcharles93/archie-core/issues/375","source_system":"github:https://github.com/samcharles93/archie-core/issues/375:375","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499114-354-e5b475b4","title":"Persistent session-memory curator (reference implementation)","description":"Ship a curator that maintains durable memory across sessions -- extracting and consolidating what is worth keeping from conversation history into the attached memory engine. The second contract consumer, and the one that exercises the curator/memory-engine binding.\n\n### Acceptance Criteria\n\nRegisters and runs through the registry with no special-casing. Writes through the memory engine contract rather than any concrete store. Extraction runs off the hot path so a chat turn never waits on it. What it wrote and why is inspectable. Honours the trigger-accounting invariant so its own writes do not reschedule it.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:54:39Z","updated_at":"2026-08-08T14:54:39Z","external_ref":"https://github.com/samcharles93/archie-core/issues/375","source_system":"github:https://github.com/samcharles93/archie-core/issues/375:375","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637499114-354-e5b475b4","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:28Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499092-353-ed4800d2","title":"Regression tests: NellDB adapter — edge cases, error recovery, concurrent access","description":"internal/nell/adapter_test.go has 7 tests covering the happy path. Missing: corrupted log recovery, concurrent Enqueue/ClaimNext, Get/List with empty store, oversized event payloads, context cancellation during operations, observer callback error isolation.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:54:34Z","updated_at":"2026-08-08T14:54:36Z","closed_at":"2026-08-08T14:54:37Z","external_ref":"https://github.com/samcharles93/archie-core/issues/374","source_system":"github:https://github.com/samcharles93/archie-core/issues/374:374","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499063-352-c2b85c1a","title":"fix(gateway): synchronize sessionTracker.active","description":"sessionTracker.active is a plain map with no mutex. go-telegram/bot dispatches each update in its own goroutine, so resolve() writes it while getActive() reads it concurrently. Backend-agnostic Go data race.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:54:29Z","updated_at":"2026-08-08T15:34:03Z","closed_at":"2026-08-08T15:34:04Z","external_ref":"https://github.com/samcharles93/archie-core/issues/373","source_system":"github:https://github.com/samcharles93/archie-core/issues/373:373","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637499041-351-b4253aa3","title":"Link pushed task branch to source issue before opening PR","description":"After the task branch is successfully pushed, call the forge branch-link operation for the source issue, then create the PR without relying on a PR comment or Closes directive to associate it. Propagate link failures and cover Gitea plus forge RPC behavior.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:54:23Z","updated_at":"2026-08-08T14:54:26Z","closed_at":"2026-08-08T14:54:27Z","external_ref":"https://github.com/samcharles93/archie-core/issues/372","source_system":"github:https://github.com/samcharles93/archie-core/issues/372:372","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -521,7 +522,7 @@
|
||||
{"_type":"issue","id":"archie-core-1786637492113-89-a9c21168","title":"Multi-agent collaboration: many agents working together","description":"Support multiple agents working in combination — not just one agent per task in isolation, but agents coordinating with each other (e.g. on the same repo, the same task, or a decomposed set of related tasks). Scope, coordination model, and conflict-avoidance strategy are all undefined — this is exploratory. No implementation approach has been decided yet.\n\n### Acceptance Criteria\n\nA PRD exists (docs/prds/) covering: what 'multiple agents working in combination' means concretely for archie-core, how agents discover/hand off work to each other, how conflicts (same repo, same worktree, shared state) are avoided or resolved, and what changes to daemon/workflow/dispatch are implied. No code changes required to close this epic — the PRD is the deliverable.","status":"open","priority":2,"issue_type":"epic","created_at":"2026-08-08T14:34:48Z","updated_at":"2026-08-08T15:35:40Z","external_ref":"https://github.com/samcharles93/archie-core/issues/110","source_system":"github:https://github.com/samcharles93/archie-core/issues/110:110","metadata":null,"labels":["epic"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637492083-88-a3a3fbff","title":"Evaluate internal/workflow architecture and APIs","description":"Perform a read-only, repository-grounded evaluation of internal/workflow. Identify correctness, definition clarity, reuse, interface ownership, coupling, testability, and maintainability improvements, with concrete evidence and prioritized recommendations.\n\n### Acceptance Criteria\n\nAssessment covers package definitions, consumers, subpackages, tests, and architecture invariants; findings are prioritized with file references and actionable recommendations; no code or git changes are made.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:44Z","updated_at":"2026-08-08T15:35:18Z","closed_at":"2026-08-08T15:35:19Z","external_ref":"https://github.com/samcharles93/archie-core/issues/109","source_system":"github:https://github.com/samcharles93/archie-core/issues/109:109","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637492059-87-2f0e575d","title":"Regression tests: internal/gate — GateContext, Finding types, Yaegi evaluation","description":"internal/gate/gate.go has zero tests. GateContext and Finding are the runtime types that custom gate scripts receive. The Yaegi evaluation path (gateeval) has tests, but the gate package itself has none. Add tests for type construction, finding validation, and integration with the gate evaluation pipeline.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:40Z","updated_at":"2026-08-08T14:34:42Z","closed_at":"2026-08-08T14:34:43Z","external_ref":"https://github.com/samcharles93/archie-core/issues/108","source_system":"github:https://github.com/samcharles93/archie-core/issues/108:108","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637492031-86-0fcaccc0","title":"Honcho memory engine as the external-provider proof","description":"Implement an engine backed by Honcho (plastic-labs/honcho, cloned at /work/clones/honcho) to prove the seam holds for a real external backend. Honcho's model is Workspace -\u003e Peer -\u003e Session -\u003e Message with asynchronously derived conclusions queried back as representations; its dual-peer convention has the assistant observe others without self-modelling. Treat this as validation of the contract, not as a dependency of the default install.\n\n### Acceptance Criteria\n\nThe engine satisfies the contract with no changes to the agent loop or gateway. Archie runs unchanged when it is not configured. Its network calls carry timeouts and its failures degrade to reported errors rather than failing a turn. Any mapping decision between Archie sessions/identities and Honcho peers/sessions is documented. Unit tests use httptest and never require a live Honcho instance, per the repo testing rule.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:34:38Z","updated_at":"2026-08-08T14:34:38Z","external_ref":"https://github.com/samcharles93/archie-core/issues/107","source_system":"github:https://github.com/samcharles93/archie-core/issues/107:107","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637492031-86-0fcaccc0","title":"Honcho memory engine as the external-provider proof","description":"Implement an engine backed by Honcho (plastic-labs/honcho, cloned at /work/clones/honcho) to prove the seam holds for a real external backend. Honcho's model is Workspace -\u003e Peer -\u003e Session -\u003e Message with asynchronously derived conclusions queried back as representations; its dual-peer convention has the assistant observe others without self-modelling. Treat this as validation of the contract, not as a dependency of the default install.\n\n### Acceptance Criteria\n\nThe engine satisfies the contract with no changes to the agent loop or gateway. Archie runs unchanged when it is not configured. Its network calls carry timeouts and its failures degrade to reported errors rather than failing a turn. Any mapping decision between Archie sessions/identities and Honcho peers/sessions is documented. Unit tests use httptest and never require a live Honcho instance, per the repo testing rule.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:34:38Z","updated_at":"2026-08-08T14:34:38Z","external_ref":"https://github.com/samcharles93/archie-core/issues/107","source_system":"github:https://github.com/samcharles93/archie-core/issues/107:107","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637492031-86-0fcaccc0","depends_on_id":"archie-core-1786637499161-356-e424e40d","type":"parent-child","created_at":"2026-09-01T15:50:34Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637492005-85-03c595e7","title":"feat(gateway): durable message deduplication via upstream IDs","description":"gateway.Message carries no upstream/channel-native message ID, so a re-delivered Telegram update appends a duplicate turn. Add Message.ID and derive the NellDB document ID from it (time-prefixed to preserve the ScanSince ID tie-break), making redelivery an idempotent LWW upsert.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:35Z","updated_at":"2026-08-08T15:34:01Z","closed_at":"2026-08-08T15:34:02Z","external_ref":"https://github.com/samcharles93/archie-core/issues/106","source_system":"github:https://github.com/samcharles93/archie-core/issues/106:106","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491986-84-5137df75","title":"feat(gateway): version the SQLite session schema before cutover","description":"OpenSQLiteSessionStore currently runs CREATE IF NOT EXISTS statements with no schema version or transactional migration record. Before production wiring, add the smallest deterministic schema-version mechanism so future upgrades cannot silently leave a partially evolved database.\n\n### Acceptance Criteria\n\nA fresh database records the current schema version; opening a supported older fixture migrates transactionally and preserves data; unsupported newer versions fail with an actionable error; migration failure leaves the prior schema/data usable; schema initialization and upgrades are covered by file-backed tests.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:34:33Z","updated_at":"2026-08-09T12:33:31Z","closed_at":"2026-08-09T12:33:32Z","external_ref":"https://github.com/samcharles93/archie-core/issues/105","source_system":"github:https://github.com/samcharles93/archie-core/issues/105:105","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491959-83-a36ef6d7","title":"test(gateway): pin SQLite FTS lifecycle and literal-query semantics","description":"Trigger-driven FTS synchronization and hostile-query escaping were manually verified but are not retained by the trimmed SQLite-specific test suite. Preserve these SQLite-specific guarantees without duplicating the shared SessionStore contract.\n\n### Acceptance Criteria\n\nTable-driven tests show FTS search/count remains correct after message deletion, ReplaceMessages, and session deletion with no cross-session leak; hostile inputs including column syntax, unmatched quotes, NEAR, wildcard, and boolean operators return literal-match behavior without syntax errors; multi-term queries pin AND semantics; tests stay SQLite-specific and focused.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:29Z","updated_at":"2026-08-08T15:33:57Z","closed_at":"2026-08-08T15:33:58Z","external_ref":"https://github.com/samcharles93/archie-core/issues/104","source_system":"github:https://github.com/samcharles93/archie-core/issues/104:104","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -530,7 +531,7 @@
|
||||
{"_type":"issue","id":"archie-core-1786637491876-80-74417b6f","title":"test(gateway): prove SQLite session rollback and reopen durability","description":"The SQLite SessionStore uses real transactions and WAL, but current tests only verify successful operations through the same open handle. Add adversarial file-backed tests that prove persisted session/message state survives close+reopen and that injected failures during Delete and ReplaceMessages roll the transaction and FTS index back to the original state.\n\n### Acceptance Criteria\n\nA file-backed store is closed and reopened with sessions, canonical MessageID, SourceID, text, sender, and timestamps intact; PRAGMA quick_check returns ok; deterministic abort triggers force mid-Delete and mid-ReplaceMessages failures; after each failure the original session, history, counts, and search results remain unchanged; focused and race tests pass.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:18Z","updated_at":"2026-08-08T15:33:59Z","closed_at":"2026-08-08T15:34:00Z","external_ref":"https://github.com/samcharles93/archie-core/issues/101","source_system":"github:https://github.com/samcharles93/archie-core/issues/101:101","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491855-79-803fa759","title":"feat(gateway): migrate canonical session/message store to SQLite","description":"docs/architecture/messaging-and-work-intake.md:107 requires the canonical session and message implementation to supply durable SQLite round trips. The current store is NellDB-backed.\n\nScope is the SETTLED parts of the messaging architecture only:\n- SQLite persistence (modernc.org/sqlite; FTS5 verified available at 3.53.3, no CGo)\n- application-generated canonical MessageID, platform IDs as correlation only (doc 88-90)\n- immutable message records (doc 112-114)\n- append-only linear sequence per conversation (doc 128)\n- real transactions for ReplaceMessages and Delete\n\nExplicitly OUT of scope: the rich typed-content / tool-call / tool-result model and delivery-attempt semantics. docs/architecture/messaging-and-work-intake.md:273-284 lists the canonical inbound message model, source-correlation model, and outbound delivery/retry/dedup semantics as DECISIONS STILL REQUIRED, so they are not specified enough to implement. That work belongs to archie-core-abg.16.\n\nDecision by Sam 2026-08-03 after the third review surfaced that the NellDB-vs-SQLite question was already settled by the architecture docs.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:34:16Z","updated_at":"2026-08-09T12:33:30Z","closed_at":"2026-08-09T12:33:31Z","external_ref":"https://github.com/samcharles93/archie-core/issues/100","source_system":"github:https://github.com/samcharles93/archie-core/issues/100:100","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491830-78-20363259","title":"fix(gateway): /compress argument dispatch is wrong in three ways","description":"session_commands.go:460-478, all CONFIRMED by execution:\n\n1. Bare /compress can never apply. isPreview is 'len(fields) == 0 || --preview || --dry-run', so /compress with no arguments takes the preview branch -- and compressPreview ends with 'Run /compress to apply'. The command tells the user to run a command that does nothing, forever. The failure-safe ReplaceMessages path added by archie-core-0ua is unreachable via the documented command; only '/compress here \u003cN\u003e' reaches it.\n\n2. /compress focus \u003ctopic\u003e silently ignores the topic. focus is validated for arity only; cfg is never modified, so it runs a full default compression and reports success. Irreversible history rewrite from a command the user believes does something narrower.\n\n3. /compress here (no N) and any unrecognised argument (/compress asdf) fall through to a full default compression rather than printing usage, unlike the sibling focus branch.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:34:12Z","updated_at":"2026-08-08T14:34:14Z","closed_at":"2026-08-08T14:34:15Z","external_ref":"https://github.com/samcharles93/archie-core/issues/99","source_system":"github:https://github.com/samcharles93/archie-core/issues/99:99","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491805-77-78d888a3","title":"Curator runtime: persistent loop with idle cooldown","description":"Implement the loop that runs a registered curator: continuous operation while work is available, backing off to a configured cooldown when there is nothing to do, waking on new input, and honouring per-curator check-in intervals.\n\n### Acceptance Criteria\n\nA curator with pending input runs without waiting for its next interval; an idle curator consumes no meaningful CPU and wakes on new input. Cooldown and check-in interval are configurable per curator. Loop timing is tested with a fake clock, not wall-clock sleeps. Shutdown is clean and bounded: an in-flight curator pass is cancelled via context and the daemon does not block indefinitely on it. A panicking curator is recovered and isolated from other curators and from the agent loop.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:34:10Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator/runtime.go implements exactly this. Verified against every acceptance-criteria clause: pending-input curators run without waiting for the interval (TestRuntimeWakeRunsPendingPassImmediately), idle curators sleep until interval (TestRuntimeIdleCuratorSleepsUntilInterval) with a configurable per-curator cooldown (TestRuntimeCooldownAfterIdlePass, TestRuntimeNextCheckInOverride), all timed via an injected fake clock not wall-clock sleep, shutdown is context-bounded (TestRuntimeShutdownCancelsInFlightPass, TestRuntimeShutdownBoundedDespiteStuckPass), and panics are recovered and isolated (TestRuntimeCheckPanicRecovered, TestRuntimePassPanicRecovered). go test ./internal/domain/curator/... -race -count=1 is fully green.","external_ref":"https://github.com/samcharles93/archie-core/issues/98","source_system":"github:https://github.com/samcharles93/archie-core/issues/98:98","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491805-77-78d888a3","title":"Curator runtime: persistent loop with idle cooldown","description":"Implement the loop that runs a registered curator: continuous operation while work is available, backing off to a configured cooldown when there is nothing to do, waking on new input, and honouring per-curator check-in intervals.\n\n### Acceptance Criteria\n\nA curator with pending input runs without waiting for its next interval; an idle curator consumes no meaningful CPU and wakes on new input. Cooldown and check-in interval are configurable per curator. Loop timing is tested with a fake clock, not wall-clock sleeps. Shutdown is clean and bounded: an in-flight curator pass is cancelled via context and the daemon does not block indefinitely on it. A panicking curator is recovered and isolated from other curators and from the agent loop.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:34:10Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator/runtime.go implements exactly this. Verified against every acceptance-criteria clause: pending-input curators run without waiting for the interval (TestRuntimeWakeRunsPendingPassImmediately), idle curators sleep until interval (TestRuntimeIdleCuratorSleepsUntilInterval) with a configurable per-curator cooldown (TestRuntimeCooldownAfterIdlePass, TestRuntimeNextCheckInOverride), all timed via an injected fake clock not wall-clock sleep, shutdown is context-bounded (TestRuntimeShutdownCancelsInFlightPass, TestRuntimeShutdownBoundedDespiteStuckPass), and panics are recovered and isolated (TestRuntimeCheckPanicRecovered, TestRuntimePassPanicRecovered). go test ./internal/domain/curator/... -race -count=1 is fully green.","external_ref":"https://github.com/samcharles93/archie-core/issues/98","source_system":"github:https://github.com/samcharles93/archie-core/issues/98:98","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637491805-77-78d888a3","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:28Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491783-76-9a0778db","title":"test(gateway): the conformance suite does not pin what it claims, and the backends already drift","description":"The commit message for 3eb00db claims the shared suite means NellDB and SQLite 'cannot drift'. Two confirmed divergences it does not catch:\n\n1. testReplaceMessagesIsFailureSafe (sessionstore_conformance_test.go:769) never induces a failure -- four happy-path cases asserting the post-condition of a SUCCESSFUL call. Swapping the write and delete loops in session_store.go:696-708 would leave it green. Only SQLite has a real abort test, and it pins rollback, the opposite failure mode. The backends genuinely differ here: SQLite is atomic (failure -\u003e clean rollback), NellDB is not (failure -\u003e duplicates). SQLite's semantics are better; the point is the suite does not know.\n\n2. NellDB's lastTS cache is never invalidated by deletes; SQLite recomputes MAX(ts). Save at 13:00, delete it, save with At=12:00 -\u003e NellDB persists 13:00:00.001Z, SQLite persists 12:00:00Z. Both pass the suite.\n\n3. GetByChannel ordering differs between backends and is unpinned (see the newest-first bead).\n\n4. applyCompress has zero test coverage: no test in the package calls it, none of its three error returns is exercised, and fakeSessionStore.ReplaceMessages is a slice assignment that cannot observe ordering or identity.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:34:08Z","updated_at":"2026-08-09T12:33:32Z","closed_at":"2026-08-09T12:33:33Z","external_ref":"https://github.com/samcharles93/archie-core/issues/97","source_system":"github:https://github.com/samcharles93/archie-core/issues/97:97","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491754-75-8e621216","title":"fix(gateway): session resolution is not newest-first, so restart resumes a stale session","description":"Found independently by two reviewers, CONFIRMED by execution.\n\nsessionTracker.resolve (session_commands.go:54-59) comments 'Find the most recent session matching this thread and bot' but returns the first element GetByChannel yields. The wired-up NellDB store returns document-ID order (listAllLocked reverses AllDocs, which sorts by document ID), i.e. UUID-lexicographic -- effectively random. LastActiveAt is populated on every record and never consulted.\n\nScenario: user runs /new 'old', chats, runs /new 'current'; daemon restarts (the tracker cache is in-memory only); the next message resolves to whichever session sorts first. Probed over 10 runs: 4 landed in the OLDER session, answering against pre-/new history. This is the user-visible lost-context class.\n\nAlso a silent backend divergence: SQLite's GetByChannel orders created_at DESC, NellDB does not, and the shared conformance suite does not pin it. Separately, SessionStore.List's doc comment (session_store.go:34-36) promises newest-first and does not deliver: saving zzz-oldest, aaa-middle, mmm-newest returns [zzz-oldest mmm-newest aaa-middle].\n\nBlocks archie-core-b4i and archie-core-gih.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:34:04Z","updated_at":"2026-08-08T14:34:06Z","closed_at":"2026-08-08T14:34:07Z","external_ref":"https://github.com/samcharles93/archie-core/issues/96","source_system":"github:https://github.com/samcharles93/archie-core/issues/96:96","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637491729-74-ee80f814","title":"Daemon: every task/event write fails with 'sdk: revision conflict' after NellDB v0.3.0 bump — counter RMW permanently broken","description":"## Symptom (observed 2026-08-05, task dev session)\n\n- Every 1m poll logs `issue queued` for the same open issue (sam/archie-core#8) again.\n- Every attempt logs `event sink insert failed: err=\"sdk: revision conflict\"` (cmd/archied/main.go:373, the event-sink goroutine calling st.InsertEvent).\n- The daemon's own task consumer logs `nats enqueue failed: err=\"sdk: revision conflict\"` (internal/daemon/daemon.go:599, calling Store.EnqueueIssue) and NAKs; JetStream redelivers to max-deliver 3 (triples of errors every ~2 min in the log).\n- The task doc is never persisted, so the enqueue dedup check always passes and the poll re-queues forever. Work never reaches the agent; dashboard shows a stuck/no-op task.\n\n## Exact cause\n\nCommit 44f914c bumped github.com/samcharles93/NellDB v0.2.5 -\u003e v0.3.0 touching only go.mod/go.sum. v0.3.0 changed the DocDB rev-cache rebuild:\n\n- v0.2.5 sdk/docdb.go reindex() indexed EVERY record by rec.ID, including Archie's meta:task_id_counter / meta:event_id_counter docs.\n- v0.3.0 sdk/docdb.go reindex() skips any id where isInternalID() is true (sdk/replicate.go:841: id[:5] == \"meta:\"), intended only for the SDK's own meta:clock/meta:vector bookkeeping but prefix-matching Archie's counter docs too.\n- v0.3.0 putIn (sdk/docdb.go:208-216) keeps the same conflict rule as v0.2.5: incomingRev != \"\" \u0026\u0026 !exists -\u003e sdk.ErrConflict.\n\nArchie's Adapter.nextCounter (internal/nell/adapter.go:957) is a Get -\u003e mutate -\u003e Put RMW that supplies the _rev returned by Get (joinDoc copies the payload, which putIn re-stamps with _rev). After any restart the counter docs are present in the log but absent from the rev cache, so exists=false and every counter Put conflicts. The counter doc can never be written again, so the cache is never repopulated — the failure is permanent for the lifetime of the DB, every process run, from the first poll.\n\nRepro (standalone, module cache, v0.3.0): open log, Put meta:task_id_counter, close, reopen, Get (rev e.g. \"1-69bfa51d...\"), Put with that rev -\u003e \"sdk: revision conflict\". First run of a fresh DB succeeds (empty-rev init Put is allowed); every subsequent process restart fails.\n\nWhy tests are green: OpenMemory/fresh-log tests initialize the counter via empty-rev Put within the same process; TestPersistenceAcrossRestart reopens a real log but only reads (TaskByIssue) and never exercises the counter RMW after reopen. The production path (restart an existing log, then enqueue) is untested.\n\n## Evidence\n\n- go.mod: github.com/samcharles93/NellDB v0.3.0 (no replace).\n- git show 44f914c --stat: go.mod, go.sum, .gitignore only; internal/nell/ untouched since.\n- Module cache diff !nell!d!b@v0.2.5 vs @v0.3.0: sdk/docdb.go reindex()/putIn(), sdk/replicate.go isInternalID.\n- /tmp/nell-repro: deterministic ErrConflict on counter RMW after reopen.\n\n## Fix direction\n\nOptions, in order of preference: (1) pin back to v0.2.5 (verified-2026-07-28 semantics) until adapter is migrated; (2) stop supplying _rev in nextCounter (drop rev from the RMW doc before Put — v0.3.0 continues the chain from the cached rev when incomingRev == \"\"), but that breaks the MVCC chain for the counter when the cache lacks it; (3) upstream: isInternalID should match exact SDK ids (meta:clock, meta:vector), not the meta: prefix, or reindex should index meta docs. Any fix must add a red test: open persisted log, EnqueueIssue, reopen, EnqueueIssue again -\u003e no conflict.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:34:01Z","updated_at":"2026-08-08T14:34:03Z","closed_at":"2026-08-08T14:34:04Z","external_ref":"https://github.com/samcharles93/archie-core/issues/95","source_system":"github:https://github.com/samcharles93/archie-core/issues/95:95","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
@@ -596,17 +597,17 @@
|
||||
{"_type":"issue","id":"archie-core-1786637490145-14-dd1e68f3","title":"Fix chat command palette keyboard interaction","description":"The chat slash-command palette remains open after the slash is removed or Escape is pressed, and its results cannot be navigated with ArrowUp/ArrowDown. Add regression tests and restore expected keyboard behavior.\n\n### Acceptance Criteria\n\nBackspace/removing the slash closes the palette; Escape closes it; ArrowUp and ArrowDown move the active option; Enter activates the highlighted option; the focused option is exposed accessibly; UI tests and project quality gates pass.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:27:05Z","updated_at":"2026-08-08T14:31:13Z","closed_at":"2026-08-08T14:31:14Z","external_ref":"https://github.com/samcharles93/archie-core/issues/35","source_system":"github:https://github.com/samcharles93/archie-core/issues/35:35","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490125-13-6dea21ca","title":"Trim the remaining agent skills to factual reference","description":"The .agents/skills/ library is 4283 lines across 12 skills. An audit found most carry useful facts buried in process ceremony. Five worst offenders were already parked at /work/apps/archie-skills-parked. Trim what remains to roughly 2200 lines total. KEEP: concrete commands, package anchors, config wiring tables, NellDB shapes, symptom-to-diagnostic indexes, known regressions, domain glossary, operational hazards, false-confidence warnings. CUT: mandated N-step procedures, handoff templates, evidence packets, acceptance checklists, ownership/deletion/rollback field lists, sibling-skill routing prose, repeated 'reject unless' instructions, dated tool snapshots. Preserve every skill's YAML frontmatter name and description. Do not delete any skill file outright.\n\n### Acceptance Criteria\n\nTotal line count of .agents/skills/*/*.md is between 1900 and 2500. All 12 skills still exist with valid frontmatter. No factual command, path, or table was lost - only procedure and ceremony. go build ./... still clean.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:27:00Z","updated_at":"2026-08-08T14:27:02Z","closed_at":"2026-08-08T14:27:03Z","external_ref":"https://github.com/samcharles93/archie-core/issues/34","source_system":"github:https://github.com/samcharles93/archie-core/issues/34:34","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490097-12-bf415ffb","title":"SOUL cannot override the prompt invariants","description":"Enforce the boundary that makes SOUL safe to hand to users: it supplies identity and style, and cannot reach the instruction hierarchy, core rules, tool inventory or env block.\n\n### Acceptance Criteria\n\nA test renders a hostile SOUL that attempts to countermand core_rules and asserts the invariant sections are still present and intact. The boundary is documented where the layering rule already lives so the next person editing the template does not quietly relax it.","status":"open","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:57Z","updated_at":"2026-08-08T14:26:57Z","external_ref":"https://github.com/samcharles93/archie-core/issues/33","source_system":"github:https://github.com/samcharles93/archie-core/issues/33:33","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490069-11-e245c170","title":"Memory engine contract and owning registry","description":"Define the typed memory engine interface and the registry that owns provider registration, discovery, and lifecycle. This is the load-bearing design step: everything else in the epic is written against this contract.\n\n### Acceptance Criteria\n\nA typed interface exposes real memory operations beyond identity metadata. An owning Registry/Manager controls registration, init, start, health, and stop, and defines failure isolation and shutdown ordering. Providers receive only a typed registrar and the host services their capability declares. internal/plugin/architecture_test.go passes with memory recognised as a conforming engine family. Stateless providers may declare explicit no-op lifecycle.","notes":"Not started as a conforming engine family. The existing internal/memory package (provider.go, lifecycle.go, backup.go, etc.) predates the domain migration: it has an ad-hoc MemoryProvider interface plus several optional hook interfaces (TurnStartHook, SessionEndHook, ...), but no owning Registry/Manager, no narrow typed registrar, and no Manifest-style declared shape -- it does not follow the strict plugin engine rule that internal/domain/curator (archie-core-1786637499933-386-83646cce, now closed) already demonstrates as the reference implementation. Confirmed via go test ./internal/plugin/... -run TestEveryEngineInterfaceHasTypedCapabilityAndFamilyOwner -v: only domain/curator.CuratorEngine, secret.Engine, and tools/provider.Engine are currently recognised conforming families -- no memory family. Use internal/domain/curator's contract.go/registry.go/registrar.go split as the template; do not extend internal/memory's MemoryProvider in place.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:54Z","updated_at":"2026-08-15T12:55:52Z","external_ref":"https://github.com/samcharles93/archie-core/issues/32","source_system":"github:https://github.com/samcharles93/archie-core/issues/32:32","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":1,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490043-10-711c450a","title":"Config surface for memory engine selection and settings","description":"Add validated configuration for choosing a memory engine and supplying its per-engine settings, following the existing config conventions in internal/config and internal/infrastructure/configuration.\n\n### Acceptance Criteria\n\nEngine choice and per-engine settings are expressible in config.toml with toml/yaml/json tags matching repo convention. Validation rejects an unknown engine name with a clear error naming the valid set, in the style of the existing forge.type validation. Defaults resolve so an absent section keeps the builtin engine. config.example.toml documents the section.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:51Z","updated_at":"2026-08-08T14:31:57Z","external_ref":"https://github.com/samcharles93/archie-core/issues/31","source_system":"github:https://github.com/samcharles93/archie-core/issues/31:31","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490069-11-e245c170","title":"Memory engine contract and owning registry","description":"Define the typed memory engine interface and the registry that owns provider registration, discovery, and lifecycle. This is the load-bearing design step: everything else in the epic is written against this contract.\n\n### Acceptance Criteria\n\nA typed interface exposes real memory operations beyond identity metadata. An owning Registry/Manager controls registration, init, start, health, and stop, and defines failure isolation and shutdown ordering. Providers receive only a typed registrar and the host services their capability declares. internal/plugin/architecture_test.go passes with memory recognised as a conforming engine family. Stateless providers may declare explicit no-op lifecycle.","notes":"Not started as a conforming engine family. The existing internal/memory package (provider.go, lifecycle.go, backup.go, etc.) predates the domain migration: it has an ad-hoc MemoryProvider interface plus several optional hook interfaces (TurnStartHook, SessionEndHook, ...), but no owning Registry/Manager, no narrow typed registrar, and no Manifest-style declared shape -- it does not follow the strict plugin engine rule that internal/domain/curator (archie-core-1786637499933-386-83646cce, now closed) already demonstrates as the reference implementation. Confirmed via go test ./internal/plugin/... -run TestEveryEngineInterfaceHasTypedCapabilityAndFamilyOwner -v: only domain/curator.CuratorEngine, secret.Engine, and tools/provider.Engine are currently recognised conforming families -- no memory family. Use internal/domain/curator's contract.go/registry.go/registrar.go split as the template; do not extend internal/memory's MemoryProvider in place.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:54Z","updated_at":"2026-08-15T12:55:52Z","external_ref":"https://github.com/samcharles93/archie-core/issues/32","source_system":"github:https://github.com/samcharles93/archie-core/issues/32:32","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637490069-11-e245c170","depends_on_id":"archie-core-1786637499161-356-e424e40d","type":"parent-child","created_at":"2026-09-01T15:50:33Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":1,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490043-10-711c450a","title":"Config surface for memory engine selection and settings","description":"Add validated configuration for choosing a memory engine and supplying its per-engine settings, following the existing config conventions in internal/config and internal/infrastructure/configuration.\n\n### Acceptance Criteria\n\nEngine choice and per-engine settings are expressible in config.toml with toml/yaml/json tags matching repo convention. Validation rejects an unknown engine name with a clear error naming the valid set, in the style of the existing forge.type validation. Defaults resolve so an absent section keeps the builtin engine. config.example.toml documents the section.","status":"open","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:51Z","updated_at":"2026-08-08T14:31:57Z","external_ref":"https://github.com/samcharles93/archie-core/issues/31","source_system":"github:https://github.com/samcharles93/archie-core/issues/31:31","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637490043-10-711c450a","depends_on_id":"archie-core-1786637499161-356-e424e40d","type":"parent-child","created_at":"2026-09-01T15:50:33Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637490011-9-d5e4f9a4","title":"Telegram: provider selector filters model choices","description":"Add an authorized native /provider inline selector. The selected provider scopes /model so only configured model references belonging to that provider are shown; e.g. openrouter/openai/gpt-5.6 must not appear under provider openai. Selecting a provider establishes a valid active model for that provider when necessary, and /status reports active provider and model.\n\n### Acceptance Criteria\n\n/provider displays configured providers as an inline keyboard and marks the active provider. Authorized callbacks switch provider; unauthorized, malformed, stale, and duplicate callbacks fail safely. /model shows only models whose first reference segment matches the selected provider and uses unambiguous labels. Switching provider leaves the chat responder on a valid model from that provider. /status includes active provider and full model reference. Menu/help/executable parity and concurrent access are tested.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:49Z","updated_at":"2026-08-18T15:53:30Z","closed_at":"2026-08-18T15:53:30Z","close_reason":"Superseded: confirmed by Sam that there is no longer a separate /provider command -- grepped the codebase and no '/provider' command exists in internal/gateway or internal/channels/telegram. /model now handles provider scoping directly.","external_ref":"https://github.com/samcharles93/archie-core/issues/30","source_system":"github:https://github.com/samcharles93/archie-core/issues/30:30","metadata":null,"labels":["enhancement","in-progress"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489985-8-13b9d8a8","title":"Document the memory engine family in ARCHITECTURE.md","description":"Record the memory engine family alongside the existing engine documentation so the contract, its lifecycle guarantees, and the trust boundary for third-party memory backends are written down where the plugin engine rule already lives.\n\n### Acceptance Criteria\n\nARCHITECTURE.md describes the memory engine contract, its owning manager, lifecycle and failure isolation, and states the trust boundary for an out-of-process or third-party memory backend. The shared CLAUDE.md/AGENTS.md source is updated if the rule surface changes.","status":"closed","priority":2,"issue_type":"task","assignee":"Sam Catlow","created_at":"2026-08-08T14:26:46Z","updated_at":"2026-08-15T13:36:39Z","started_at":"2026-08-15T13:25:55Z","closed_at":"2026-08-15T13:36:39Z","close_reason":"Documented the memory engine contract, owning manager, lifecycle/failure isolation, and third-party trust boundary in ARCHITECTURE.md; focused memory tests pass.","external_ref":"https://github.com/samcharles93/archie-core/issues/29","source_system":"github:https://github.com/samcharles93/archie-core/issues/29:29","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489959-7-6c8cdec2","title":"fix(telegram): honour retry_after on 429 instead of the library's fixed backoff","description":"go-telegram/bot v1.22.0 get_updates.go discards TooManyRequestsError.RetryAfter and applies its own 100ms-doubling backoff capped at 5s. Against retry_after:5 that is ~12 further requests inside the silence window, which escalated a brief 429 on carina into sustained 429s, 502s and request timeouts over several minutes.\n\nFixed in internal/channels/telegram/ratelimit.go: WithErrorsHandler is invoked synchronously from the polling loop before the library applies its own backoff, so blocking there honours the interval without patching the dependency. errors.As unwraps the library's fmt.Errorf wrap; capped at maxRetryAfter=60s; releases immediately on context cancellation so /restart during a 429 does not block.\n\nUpstream report drafted for https://github.com/go-telegram/bot/issues -- suggested fix is to prefer RetryAfter over incErrTimeout in the error branch.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:41Z","updated_at":"2026-08-08T14:26:44Z","closed_at":"2026-08-08T14:26:45Z","external_ref":"https://github.com/samcharles93/archie-core/issues/28","source_system":"github:https://github.com/samcharles93/archie-core/issues/28:28","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489932-6-1f803e02","title":"Curator observability surface","description":"Give curator activity visibility: which curators are registered, when each last ran, what it changed, and why. Background agents mutating memory unobserved is the failure mode this epic most needs to avoid.\n\n### Acceptance Criteria\n\nRegistered curators, their schedule state, and their recent activity are inspectable at runtime. Each recorded action carries enough context to answer why it happened. Surfaced through the existing observability paths rather than a bespoke channel.","notes":"Half-built: the data this needs already exists but nothing surfaces it. internal/domain/curator/registry.go has Health(ctx) map[string]Health and Names() []string; every pass emits an Action (At/Type/Detail/Reason) and a run event through the curator's EventSink. daemon.Daemon.Curators *curator.Registry (internal/daemon/daemon.go:107-110) holds the live registry. But grep confirms no webui/API/dashboard code reads any of it -- internal/webui has zero curator-aware endpoints, ui/src has no curator surface. This issue is 'wire the existing Registry.Health/Names + action history into the existing observability paths (health endpoint / dashboard)', not 'build the data model from scratch'.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:38Z","updated_at":"2026-08-27T14:31:41Z","closed_at":"2026-08-27T14:31:41Z","close_reason":"Wired GET /api/curators (names, health, last-run, recent actions) backed by a new bounded in-memory activity tracker in internal/domain/curator, plus a Curators dashboard page at ui/src/curators/.","external_ref":"https://github.com/samcharles93/archie-core/issues/27","source_system":"github:https://github.com/samcharles93/archie-core/issues/27:27","metadata":null,"labels":["enhancement"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489932-6-1f803e02","title":"Curator observability surface","description":"Give curator activity visibility: which curators are registered, when each last ran, what it changed, and why. Background agents mutating memory unobserved is the failure mode this epic most needs to avoid.\n\n### Acceptance Criteria\n\nRegistered curators, their schedule state, and their recent activity are inspectable at runtime. Each recorded action carries enough context to answer why it happened. Surfaced through the existing observability paths rather than a bespoke channel.","notes":"Half-built: the data this needs already exists but nothing surfaces it. internal/domain/curator/registry.go has Health(ctx) map[string]Health and Names() []string; every pass emits an Action (At/Type/Detail/Reason) and a run event through the curator's EventSink. daemon.Daemon.Curators *curator.Registry (internal/daemon/daemon.go:107-110) holds the live registry. But grep confirms no webui/API/dashboard code reads any of it -- internal/webui has zero curator-aware endpoints, ui/src has no curator surface. This issue is 'wire the existing Registry.Health/Names + action history into the existing observability paths (health endpoint / dashboard)', not 'build the data model from scratch'.","status":"closed","priority":2,"issue_type":"feature","created_at":"2026-08-08T14:26:38Z","updated_at":"2026-08-27T14:31:41Z","closed_at":"2026-08-27T14:31:41Z","close_reason":"Wired GET /api/curators (names, health, last-run, recent actions) backed by a new bounded in-memory activity tracker in internal/domain/curator, plus a Curators dashboard page at ui/src/curators/.","external_ref":"https://github.com/samcharles93/archie-core/issues/27","source_system":"github:https://github.com/samcharles93/archie-core/issues/27:27","metadata":null,"labels":["enhancement"],"dependencies":[{"issue_id":"archie-core-1786637489932-6-1f803e02","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:29Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489903-5-fde58498","title":"CLI: technical design interview and PRD for command surface","description":"Run a technical design interview with Sam before making any CLI implementation decisions. Produce a reviewed PRD/specification for the user-facing archie CLI: goals, personas, installation/distribution, configuration and state model, commands/subcommands/flags, output and machine-readable contracts, auth/host targeting, update lifecycle, error semantics, compatibility/migration, and acceptance tests. The eventual implementation must use urfave/cli/v3; do not use or introduce urfave/cli/v2. No CLI code is in scope until Sam approves the resulting PRD.\n\n### Acceptance Criteria\n\nInterview captures Sam's intended CLI workflows and non-goals; PRD defines the complete command surface and state/transport ownership without assumptions; PRD specifies urfave/cli/v3; Sam explicitly approves the PRD before implementation work is created.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:34Z","updated_at":"2026-08-08T14:26:36Z","closed_at":"2026-08-08T14:26:37Z","external_ref":"https://github.com/samcharles93/archie-core/issues/26","source_system":"github:https://github.com/samcharles93/archie-core/issues/26:26","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489876-4-35aced49","title":"feat(config): optional file logging","description":"archied writes slog JSON to stdout only; under systemd journald is the sole copy, and running by hand leaves nothing. Add a log_file config field so there is a durable copy independent of journald. Prerequisite for the webui logs view showing anything from before the dashboard connected.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:29Z","updated_at":"2026-08-08T14:26:31Z","closed_at":"2026-08-08T14:26:32Z","external_ref":"https://github.com/samcharles93/archie-core/issues/25","source_system":"github:https://github.com/samcharles93/archie-core/issues/25:25","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489857-3-24abe68c","title":"fix(gateway): make /compress failure-safe","description":"applyCompress swallows MessageCount and DeleteRecentMessages errors (allCount, _ := and _, _ =) then deletes all messages BEFORE saving the compressed replacement. A crash or error between the two loses the session's entire history unrecoverably. Add a store-level ReplaceMessages that saves first and deletes the previously-listed IDs after, so the failure mode is recoverable duplicates rather than total loss.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:24Z","updated_at":"2026-08-08T15:33:31Z","closed_at":"2026-08-08T15:33:32Z","external_ref":"https://github.com/samcharles93/archie-core/issues/24","source_system":"github:https://github.com/samcharles93/archie-core/issues/24:24","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489834-2-109e7cab","title":"[BUG] /model \u003cname\u003e asked the router for a doubled model name","description":"The nestif extraction in 651de0a pulled --provider parsing out of parseModelRest into extractProviderFlag, which took the modelParts slice by pointer and appended every non-provider field to it. The caller's own loop then appended the same fields again, so every field of the model name was duplicated: '/model gpt-5.6' resolved to 'gpt-5.6 gpt-5.6' and SetActiveModel rejected it as an unknown model. '/model gpt-5.6 --provider openai' was worse -- the space-separated provider value was no longer consumed with its flag, giving 'gpt-5.6 gpt-5.6 openai'. Direct model switching over Telegram was broken for anything but a bare --refresh or a provider-only switch. Found while auditing 6d4e0be for archie-core-ifi: the naked return in the same function could not be regression-tested until the function produced correct values.\n\n### Design\n\nTwo ways to fix the double append were available. Rejected: keep the helper\nsignature and delete the caller's default-branch append. That restores correct\noutput but leaves the shape intact -- a helper holding a pointer to the\ncaller's accumulator, with two loops over the same fields -- so the next person\nwho adds a field type has the same trap waiting. It also does not fix the\nspace-separated provider value, because the helper returns early on the first\n--provider and never sees the fields after it.\n\nChosen: splitProviderFlag returns (provider, remaining) and is the sole\naccumulator. One loop owns the fields, and there is no shared slice to append\nto twice. The index loop is kept deliberately -- i++ after reading a\nspace-separated value is what removes it with its flag, and converting that to\na range loop is precisely what 651de0a did wrong.\n\nLast --provider wins if repeated, matching the pre-651de0a behaviour rather\nthan the buggy helper's first-wins. Not worth a flag: /model does not document\nrepeating it.\n\n### Acceptance Criteria\n\nparseModelRest returns the model name exactly once for bare names, scope flags, and both --provider spellings, with a regression test asserting field count rather than string equality.","status":"closed","priority":2,"issue_type":"bug","created_at":"2026-08-08T14:26:19Z","updated_at":"2026-08-08T15:34:07Z","closed_at":"2026-08-08T15:34:08Z","external_ref":"https://github.com/samcharles93/archie-core/issues/23","source_system":"github:https://github.com/samcharles93/archie-core/issues/23:23","metadata":null,"labels":["bug"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489768-1-2fd8288f","title":"Trigger accounting: curator output must not feed its own trigger","description":"Establish the invariant that work a curator produces does not count toward the condition that schedules the next curator run. The scheduler must explicitly exclude derived curator output from the observation threshold: counting it would inflate the threshold and create a feedback loop, so a curator that writes memories must not thereby schedule itself again.\n\n### Acceptance Criteria\n\nDerived curator output is distinguishable from primary input at the point the trigger is evaluated. A curator that writes memories does not thereby schedule itself again. There is a regression test that would fail if derived output were counted, with the feedback-loop rationale recorded in the test so the guard is not removed as redundant.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:17Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator/wake.go WakeOnPrimaryInput forwards only explicitly-passed inputKinds; curator-produced kinds (curator_run, curator_action, curator_error) are structurally excluded, per its own doc comment naming this exact invariant. internal/domain/curator/wake_test.go:11-21 TestWakeOnPrimaryInputForwardsOnlyPrimaryKinds carries the required feedback-loop rationale verbatim ('if derived curator output counted... a curator that writes memories would schedule itself again... This test fails if a derived kind is added to the forwarded set or the filter is removed as redundant') and asserts passCalls/checkCalls stay at 1 after publishing derived-kind events. All three acceptance-criteria clauses satisfied exactly.","external_ref":"https://github.com/samcharles93/archie-core/issues/22","source_system":"github:https://github.com/samcharles93/archie-core/issues/22:22","metadata":null,"labels":["task"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1786637489768-1-2fd8288f","title":"Trigger accounting: curator output must not feed its own trigger","description":"Establish the invariant that work a curator produces does not count toward the condition that schedules the next curator run. The scheduler must explicitly exclude derived curator output from the observation threshold: counting it would inflate the threshold and create a feedback loop, so a curator that writes memories must not thereby schedule itself again.\n\n### Acceptance Criteria\n\nDerived curator output is distinguishable from primary input at the point the trigger is evaluated. A curator that writes memories does not thereby schedule itself again. There is a regression test that would fail if derived output were counted, with the feedback-loop rationale recorded in the test so the guard is not removed as redundant.","status":"closed","priority":2,"issue_type":"task","created_at":"2026-08-08T14:26:17Z","updated_at":"2026-08-15T12:55:20Z","closed_at":"2026-08-15T12:55:20Z","close_reason":"Complete: internal/domain/curator/wake.go WakeOnPrimaryInput forwards only explicitly-passed inputKinds; curator-produced kinds (curator_run, curator_action, curator_error) are structurally excluded, per its own doc comment naming this exact invariant. internal/domain/curator/wake_test.go:11-21 TestWakeOnPrimaryInputForwardsOnlyPrimaryKinds carries the required feedback-loop rationale verbatim ('if derived curator output counted... a curator that writes memories would schedule itself again... This test fails if a derived kind is added to the forwarded set or the filter is removed as redundant') and asserts passCalls/checkCalls stay at 1 after publishing derived-kind events. All three acceptance-criteria clauses satisfied exactly.","external_ref":"https://github.com/samcharles93/archie-core/issues/22","source_system":"github:https://github.com/samcharles93/archie-core/issues/22:22","metadata":null,"labels":["task"],"dependencies":[{"issue_id":"archie-core-1786637489768-1-2fd8288f","depends_on_id":"archie-core-1786637500725-414-9d66a52b","type":"parent-child","created_at":"2026-09-01T15:50:29Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-jkmu","title":"send_file confinement is symlink-blind and re-opens the file after checking it","description":"internal/tools/builtin/confinement.go ResolveReadable applies exactly the read tool's policy (resolvePath + isReadConfined), which is purely lexical: filepath.Abs/Rel, never EvalSymlinks. A symlink inside the workspace pointing at /etc/shadow passes the check, and the telegram sender then opens att.Path with no confinement check of its own, in a different goroutine after the turn. There is also a real TOCTOU window -- prepare stats/opens/closes, SendMedia re-stats and re-opens later.\n\nThe policy is unchanged from the read tool, so this is not a new hole; what changed is the consequence. The read tool caps at 5MB into the operator's own context; send_file ships up to 50MB OUT to a chat. The shell tool has never been confined either, so a model that wants the bytes already has them.\n\nDecide: either resolve symlinks in ResolveReadable (which changes read-tool behaviour too, and may break legitimate symlinked workspaces), or document the caveat explicitly and stop claiming 'sending cannot reach what reading cannot' without it.\n\nFound by adversarial review of 9a45fbb.\n\n## Acceptance Criteria\nEither the symlink case is closed at the resolver, or the limitation is written down where the claim is made (confinement.go doc comment and sendfile package doc).","status":"open","priority":3,"issue_type":"task","owner":"sam@catlow.tech","created_at":"2026-08-24T15:54:43Z","created_by":"Sam Catlow","updated_at":"2026-08-24T19:49:55Z","external_ref":"https://github.com/samcharles93/archie-core/issues/647","dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-1ysw.8","title":"Honor cancellation before local worktree commits","description":"CommitAll discards its context and can stage and commit after task cancellation.\n\nConfirmed still present on main: `Manager.CommitAll` (internal/worktree/worktree.go:387) takes `_ context.Context` and never reads it. Between entry and `wt.Commit` it performs `AddWithOptions{All: true}` and `wt.Commit`, neither of which is context-aware in go-git, so a task cancelled just before the call still mutates the index and HEAD.\n\n## Steps to Reproduce\n\n1. Prepare a task worktree with uncommitted changes.\n2. Create an already-cancelled context: `ctx, cancel := context.WithCancel(context.Background()); cancel()`.\n3. Call `Manager.CommitAll(ctx, dir, \"msg\")`.\n4. Observe it returns `(true, nil)`: the index was staged and a new commit is on HEAD despite the context being cancelled before the call.\n5. In production this leaves a commit on the task branch for work the operator cancelled, which the subsequent push/PR path can then publish.\n\n## Acceptance Criteria\n\nA pre-canceled context causes no index or HEAD mutation; cancellation is checked between non-context-aware go-git phases where possible, with regression tests.\n","acceptance_criteria":"A pre-canceled context causes no index or HEAD mutation; cancellation is checked between non-context-aware go-git phases where possible, with regression tests.","status":"closed","priority":3,"issue_type":"bug","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-22T11:01:48Z","created_by":"Sam Catlow","updated_at":"2026-08-24T08:15:48Z","started_at":"2026-08-24T07:42:10Z","closed_at":"2026-08-24T08:15:48Z","close_reason":"Fixed: CommitAll now takes a live context and checks ctx.Err() before open, worktree, staging, status and commit (worktree.go:398-430), so a pre-cancelled context mutates neither index nor HEAD. Proven red against pre-fix code: 'CommitAll() error = nil, want context canceled' for both a cancelled context and an expired deadline. task check green; fresh adversarial review zero findings.","external_ref":"https://github.com/samcharles93/archie-core/issues/603","dependencies":[{"issue_id":"archie-core-1ysw.8","depends_on_id":"archie-core-1ysw","type":"parent-child","created_at":"2026-08-22T21:01:47Z","created_by":"Sam Catlow","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
{"_type":"issue","id":"archie-core-0y4v","title":"agent: markWorktreeSafe checks existence, not mount-ness, so a stale dir writes to the operator's gitconfig","description":"markWorktreeSafe gates on os.Stat (existence and dir-ness), not mount-ness, despite the doc comment claiming a 'mount' check so host-mode agents don't write to the operator's own ~/.gitconfig for a path that was never mounted (commit 6b46ea1's stated purpose). A directory that merely exists at /data/worktree (a stale leftover) satisfies the guard and appends safe.directory=/data/worktree to the operator's gitconfig. Low severity — the entry is inert unless a repo exists at that exact path — but the guard's behavior does not match its documented contract, and the exists-but-not-a-mount case is untested.\n\nFailure: a host-mode agent (local-ollama-standalone.toml) on a machine with a leftover /data directory: an extra safe.directory line appears in the operator's ~/.gitconfig, exactly what the guard claims to prevent.","design":"Either implement a real mount check (e.g. /proc/self/mountinfo on Linux) or correct the doc comment to say the guard is existence-based and accept the (inert) entry. Decide which contract is intended.","acceptance_criteria":"Behavior matches the documented contract; the exists-but-not-mounted case is covered by a test, or the doc is corrected.","status":"closed","priority":3,"issue_type":"bug","assignee":"Sam Catlow","owner":"sam@catlow.tech","created_at":"2026-08-14T14:43:38Z","created_by":"Sam Catlow","updated_at":"2026-08-14T15:48:45Z","started_at":"2026-08-14T15:47:55Z","closed_at":"2026-08-14T15:48:45Z","close_reason":"Corrected markSafe's doc comment to match its actual existence-based (not mount-based) check; pinned the exists-but-unmounted case with a new test rather than adding platform-specific mount detection for an inert failure mode. Committed f1d74ab.","external_ref":"https://github.com/samcharles93/archie-core/issues/494","labels":["agent","code-review"],"dependency_count":0,"dependent_count":0,"comment_count":0}
|
||||
|
||||
Reference in new issue
Block a user