Seventh slice of docs/prds/eda-playbook-engine.md (t2db.15), consuming
t2db.14's expr package: the rich trigger+actions playbook document and
the coordinator that dispatches it.
- internal/domain/eda/playbook: Load(dir) reads *.yaml/*.yml playbook
documents (trigger + exactly one workflow-kind action + optional CEL
when), validating the HARD BOUNDARY at load: a playbook with 2+ actions
or a non-workflow action position is a reported load failure --
multi-action and Module/Channel/Forge actions stay blocked on the
unresolved execution-time gaps. Any malformed YAML, invalid trigger,
missing workflow name, or when compile failure fails the WHOLE load
(reject-at-load, same shape as routing collisions); a missing dir is an
empty store. Trigger reuses the existing workintake kind/label
vocabulary (no second label-matching mechanism). Dispatch matches
trigger, evaluates when (CEL, event context with actions: {} -- no
prior actions in single-action playbooks), and returns the named
workflow from the registry via the same lookup Route() uses -- no
duplicated dispatch logic. when eval error follows the resolved doc's
J3: false, skip, caller logs.
- Store.Dispatch(reg, DispatchInput) is the coordinator entrypoint; input
carries Labels/Kind/Event. Production task intake wiring
(daemon.go/pollNATS/publishTask) is explicitly NOT in this ticket -- a
follow-up; proven end to end via direct tests plus the shipped example
(examples/eda-playbooks/bug-tdd.yaml) loaded and dispatched through the
real path.
- Config: EDAPlaybookDir (eda_playbook_dir), documented as DISTINCT from
PlaybookDirs (flat binding files) -- a reader cannot confuse the two.
- Bootstrap: playbook.Load wired in loadWorkflows, log-and-abort on
failure matching every prior slice; nonexistent dir is an empty store.
- workflow.Route(), workflowForLabels, and the t2db.9-.12 binding loaders
are UNTOUCHED (verified: their full test suites pass, zero regressions).
Also carries small lint cleanups on t2db.13/.14 code flagged by the gate
run: expr Eval no longer type-asserts to ref.Val (Eval returns ref.Val
directly in v0.32.0; cel-go's own Eval wraps evaluation in recover --
verified in source); module decodeLogArgs extracted to clear the nestif
finding. golangci-lint: 0 issues.
Tests (10): single-action round trip, when-false skip, 2-action load
failure, non-workflow position load failure, when compile failure drops
playbook, empty trigger load failure, trigger mismatch skip, no-when
match, one-invalid-file fails whole load, missing dir empty store, plus
the shipped example load+dispatch. task check clean (113/113 UI, exit 0).
Fifth slice of docs/prds/eda-playbook-engine.md (t2db.13), implementing
docs/prds/module-position.md's recommended first slice: prove schema ->
go:generate -> Yaegi load -> ModuleRegistry end to end with a
side-effect-free action kind, before any kind with a real external effect
(notify, image-gen, forge writes) -- those stay blocked on the parent
doc's unresolved execution-time gaps (idempotency, mid-run failure).
- internal/domain/eda/module/log: hand-written schema source (Args/Result/
SchemaVersion per module-position.md) + logextract generated symbol
table via go:generate yaegi extract (same tooling as wfextract/
gateextract; scoped to ./internal/domain/eda/module/...).
- internal/domain/eda/module: ModuleRegistry with Register(kind, dir) /
Invoke(ctx, kind, rawArgs). No generic Module interface with an any
payload: each kind is its own tiny package with its own generated
contract; the registry's internal storage is the only place erasure
appears (exactly as inside wfextract). Invoke decodes rawArgs strictly
into the kind's Args before calling the interpreted function -- an
unknown key or wrong type is a reported failure, not a silent
zero-value fill (the 'schema defines the accepted message' rule).
Generalizes gateeval's fixed-signature Yaegi pattern via
yaegiutil.New/Resolve/Safe -- no new interpreter-construction or
panic-recovery mechanism.
- Config: ModuleDir mirroring PluginDir's shape and doc comment
(operator-trusted directory of Yaegi module implementations, one .go
per kind; empty means none loaded). Wired at daemon startup in
bootstrap.go alongside PlaybookDirs loading -- log-and-abort on load
failure (a broken module is a startup failure, not degrade-and-skip).
- examples/modules/log.go: operator-facing example (//go:build ignore,
mirroring examples/secret-engines/) proving the shipped contract.
Tests: registry (register/invoke, unknown kind, missing file, strict
arg-decode shape failures, malformed file, panic recovery), logextract
smoke test, and an example-load test that runs the shipped
examples/modules/log.go through the real registry (not a test fixture).
task check clean (113/113 UI, exit 0).
Playbook YAML action dispatch is deliberately out of this slice: a
minimal 'run this one module action, no chaining' dispatch still requires
naming how args values reach the action (parent doc open question 1 --
data-flow/interpolation syntax) and what the result feeds, both
unresolved by design. Dispatch is a follow-up ticket that consumes this
registry as-is; recorded in module-position.md.
Third slice of docs/prds/eda-playbook-engine.md (t2db.11), completing open
question 2: instead of a single playbook dir, the config gains a LIST of
directories (playbook_dirs) of *.yaml/*.yml binding files, loaded at
startup as an additional input to workflow_routing_file and
workflow_labels_file, which remain supported unchanged.
- LoadPlaybookDirs(dirs []string) reads every yaml file across the
configured directories and merges kind + arbitrary-label bindings.
- This is what finally exercises the design doc's drop-and-report
collision rule: a single file cannot collide with itself, so the
cross-source path (two files in one dir, two dirs, or a dir plus a
single-file field) had never run in production. A binding key declared
by more than one source is a reported load failure, not a silent
pick-one by source order or version.
- Router-level collision handling: MergeKindWorkflows/MergeLabelWorkflows
fail on a key bound by both the single-file fields and the playbook
dirs, applied at boot.
- Config: PlaybookDirs []string, empty by default (no behavior change for
existing configs). Repo-scoping (typing dirs to Config.Repos owner/name)
is explicitly deferred to a separate later decision per ticket;
org/tenant keying is out of scope.
- PRD: records the list shape per Sam's correction (single dir would
contradict multiple independently-maintained playbook sources).
Second slice of docs/prds/eda-playbook-engine.md, complementing the
kind->workflow layer (WorkflowRoutingFile): the label vocabulary itself
becomes playbook data. A label outside the closed bug/feature/bootstrap
kind set (e.g. security) now dispatches to a registered workflow from an
optional WorkflowLabelsFile, instead of falling through to triage.
- LabelWorkflows + LoadLabelWorkflowsYAML/SetLabelWorkflows in
internal/domain/workflow/routing.go, mirroring the kind layer's
once-at-startup lifecycle. The closed Kind/NATS-subject set is
untouched; the map only extends binding authority for labels the kind
layer does not own.
- Collision rule per design doc: a label owned by the kind set, an empty
label, an empty workflow name, or a duplicate binding is a reported
load failure (dropped, logged, error returned to the caller) -- the
schema defines what is accepted; nothing is silently arbitrated by
schema/version.
- Route() precedence: explicit t.Workflow -> arbitrary-label binding ->
kind binding -> triage -> implement -> default.
- Config: WorkflowLabelsFile (workflow_labels_file), wired at boot
alongside WorkflowRoutingFile; documented in config.example.toml.
- PRD: records this slice's decisions (single-file shape, collision rule,
precedence) so later slices stay anchored.
binding.Binding.Secret was persisted as plaintext in the bindings.secret
column, so anyone with the SQLite file saw sender-side HMAC secrets. Seal
it with AES-256-GCM transparently inside the store: encrypt-on-write,
decrypt-on-read.
- cipher: AES-256-GCM, 12-byte random nonce, SHA-256-normalised 32-byte
key resolved through the secret registry ([bindings].encryption_key).
- envelope 'arcie-binding:v1:<fingerprint>:<b64(nonce||ct||tag)>' so a
future cipher/KDF change is a version bump and old rows stay readable.
- rotation via previous_encryption_keys; key identity is the material's
SHA-256 fingerprint embedded in each row.
- legacy plaintext rows written before a key was configured stay readable;
encryption is not retroactive.
- nil cipher (no key configured) preserves the pre-t2db.7 plaintext path.
Deferred: captured_events.body redaction upgrade (separate concern, see
docs/prds/binding-secret-encryption.md).
Refs archie-core-t2db.7
RepoView.ReviewEnabled was previously absent from GET /api/config
entirely, not just unrendered; add it alongside AllowConcurrent and
MaxRetries. The repositories table gains Concurrent/Max retries/
Self-review columns as read-only pills -- this is the concrete
discoverability fix for "why isn't a second agent running": that
config was already reachable from the API, just never shown.
config.example.toml documents review_enabled, which had no example
entry before this.
Full inline editing of per-repo fields is out of scope here: the
dashboard's PATCH /api/config path has no array-index support for
updating one element of the Repos slice, which is separate,
real scope -- filed as archie-core-b6ew.6.
Closes archie-core-b6ew.4
Adds internal/domain/image (Provider/Registry per the plugin engine
family shape, but no Lifecycle -- a provider is call-scoped, not
long-running) and config.ImageConfig with the same explicit
enabled=true cost gate MinimaxConfig already uses, since a hosted
provider spends real API credits and a local one can saturate the
host GPU. Design doc at docs/prds/image-capability-contract.md per
CLAUDE.md's rule for undesigned capabilities.
No provider implementation lands here -- that's the epic's next
child bead. Also re-links all 7 epic children that were unlinked
from archie-core-1786748942120-1-6636e629 (same orphaning pattern
found in the curator/memory epics earlier this session).
Closes archie-core-1786637499161-356-e424e40d.1.
setupMemory's own doc comment claimed external providers are added via
RegisterExternal when cfg.Memory.Provider is set -- no such call
existed. Investigated wiring it for real: there is no external
MemoryProvider implementation anywhere in the repo to construct and
register (only internal/memory/builtin.Provider exists), so there is
nothing to route memory.provider's value to yet. Wiring the call would
be dead plumbing to nothing, not a fix.
The actual defect is that setting memory.provider has always silently
done nothing -- an operator who set it reasonably believed they
configured something. validateMemory now rejects it outright with a
clear reason, turning a silent no-op into a loud one at config load,
same failure-loudly posture as every other unimplemented-value check in
this file. Drop the rejection once a real external provider (or the new
domain/memory engine family's Honcho proof, archie-core-1786637492031)
gives the field an effect again.
Updated setupMemory's doc comment and config.example.toml to match --
the provider_config example is removed too, since it has nothing left
to configure.
Part of archie-core-1786637490043 (config surface for memory engine
selection and settings).
config.MemoryConfig gets an Engine field selecting among the
domain/memory engine family, plus toml tags on every field -- Provider,
ProviderConfig, and SessionTTL had none before, yaml/json only, which
didn't match repo convention.
Engine and Provider are kept as two separate fields rather than one:
they select between two different mechanisms (the new engine family vs.
the legacy external-provider path, archie-core-1786637499161-356-e424e40d.1),
and conflating them would mean a value meant for one silently getting
read by the other once both are wired.
applyMemoryDefaults resolves an absent engine to "builtin" so
unconfigured and explicitly-builtin are the same state everywhere after
defaulting. validateMemory rejects an unrecognised engine name in the
same style as forge.type's existing validation -- named constant list,
"%w: field %q (want %s)" shape, wired into the same validate() chain.
config.example.toml documents the new field.
Not yet wired to anything: nothing reads cfg.Memory.Engine at daemon
boot yet to actually select and start an engine through
domain/memory.Registry. That's bootstrap wiring, tracked as the
remaining half of archie-core-1786637499480 (ship the builtin store as
the default engine).
Origin validation compared the request's Origin against the daemon's own
direct connection: scheme from r.TLS, host from r.Host. Behind a TLS-
terminating reverse proxy the browser sends Origin: https://host while the
daemon sees a plain HTTP connection, so every dashboard mutation was
refused as cross-origin.
Taking X-Forwarded-Proto/-Host unconditionally would be worse than the bug:
any client can set them, so an exposed daemon would let an attacker forge
the scheme half of the CSRF check. They are honoured only when
web.trust_forwarded_headers is explicitly enabled, which is off by default
and documented as safe only behind a proxy that overwrites or strips
untrusted forwarded headers.
authorizeTaskMutation becomes a method because the decision now depends on
server configuration; its three call sites move with it.
Autonomous workflows now always hand a complete task to a scoped archie-agent container. Embedded NATS is the default and external NATS changes only broker placement; legacy agent and containers.enabled inputs decode for migration but cannot select an execution path. The daemon flushes task-event subscriptions before a worker can publish, and runtime updates snapshot and restore the configured worker image.
Autonomous workflows now have exactly one production boundary: a task-scoped archie-agent receives a full-task core-NATS request and runs the worker-local stage loop. The legacy [agent] section remains decode-only for migration; it no longer selects runtime behavior.
Render completed ToolCallEvents as Telegram-safe streaming blocks
with compact name/status and a short cleaned preview. Summarize
file contents and listings instead of dumping source or quoted
JSON. Collapse repeated legacy turn-budget failures.
Remove max_tokens turn gating and the 200000 turn_budget_chars
default so absent or commented config cannot restore an aggregate
cap. Keep per-result max_result_chars and spill.
Confine find/grep discovery to the active workspace/worktree.
Refs #604
The check-side of the reference update adapter never existed:
scripts/ had archie-update-install and archie-update-watchdog but no
archie-update-check, so any deployment that configured
chat.telegram.update_check_command pointing at the documented
"/usr/local/bin/archie-update-check" path had nothing there to run --
/update would report "not configured" regardless of what was actually
set. Add it: read-only, reports the daemon/agent components via git
describe against the local checkout and origin/main (no fetch-and-pull
the way install does), using the same "daemon"/"agent" component IDs
archie-update-watchdog's report and releaseupdate.ComponentDaemon/
ComponentAgent already use, so Service.Enrich and Report.Verify (this
session's earlier work) recognise it automatically.
Also fix archie-update-install's archie-agent build: it stamped
archied's ldflags but built archie-agent with none at all, so even a
native (non-container) archie-agent install reported an unstamped
"dev" build with installtype.Unknown.
config.example.toml gained every section that existed in the Go schema
but wasn't documented anywhere, not even commented: skills_dir,
plugin_dir, [chat.email], [chat].webhook_addr, [log], [memory] (noted
as currently unwired -- RegisterExternal is declared but never called
from cmd/archied, so the section has no effect yet), [[identities]]
(pointer to docs/architecture/identity.md rather than duplicating it),
and [[repos]].max_retries. tools.tool_policy.parallel_execution was
deliberately left undocumented: it's declared on ToolPolicy but never
read anywhere, genuinely dead rather than just underdocumented.
drainChatStream now detects a tools.MultimodalResult in a tool
result's Output and reports each URL via turn.Media(...), alongside
the existing turn.ToolCall(...) narration -- the missing link between
a tool producing media and a channel delivering it.
MinimaxConfig registers the tool only once an API key actually
resolves (registerMinimaxTool), matching web_fetch's "never advertise
a broken capability" pattern. Off by default, unlike web_fetch: this
spends real API credits per call.
Add [forge] intake = "poll" (default) | "webhook" | "both", and a forge
webhook receiver (internal/forge/webhook) that verifies a GitHub webhook's
HMAC signature, decodes the "issues" event, and publishes a matched issue
through the same path the poller uses -- so a labelled or assigned issue
starts a task the moment it happens instead of waiting up to poll_interval
(bead archie-core-7d5u.4).
The dispatch predicate (workintake.MatchesDispatch/RequiresLabel) is now a
single shared definition consumed by both the poller and the webhook
receiver, so the two cannot drift -- including the empty-label guard from
e19ff3a (an empty label must never match every open issue). Assignee
matching is case-insensitive to agree with AssignedIssues' server-side
GitHub filter; label matching stays case-sensitive, since GitHub labels
are case-distinct.
daemon.publishTask is exported as PublishTask, the single enqueue path
now shared by the poller and the webhook intake.
Two failure modes fixed during adversarial self-review before this landed:
a webhook secret that fails to resolve at runtime now degrades (logs and
skips the receiver) instead of aborting the whole daemon -- under
intake="both" the poll exists specifically as a backstop for this exact
misconfiguration, so crashing boot over it defeated its own purpose;
multi-identity deployments ([[identities]]) are explicitly unsupported for
now (webhook intake refuses to start, logs why, and leaves each identity's
own poll loop unaffected) rather than silently matching issues against the
wrong identity's dispatch rules.
Wire github.com/nats-io/nats-server/v2 into production startup so a
single-process deployment gets reaction delivery and task distribution
over an in-process server rather than a separate one
(docs/prds/embedded-nats.md).
Meaning change: [nats].url empty used to mean "no NATS, use the SQLite
ClaimNext flow"; it now means "start an embedded NATS server". The old
behaviour is the explicit opt-out [nats].mode = "off". mode resolves from
url when unset (url set -> external, url empty -> embedded), and
agent.mode="nats" / [containers] now require mode="external" (an embedded
loopback server cannot be reached by agent workers or containers).
The daemon owns the embedded server's lifecycle: StartEmbedded before
dialing the client, Shutdown after the client closes (LIFO cleanups). The
embedded server binds loopback on a random port, enables JetStream in a
store dir under the data directory, and leaves its logger nil so a startup
failure surfaces as a Connect error rather than an os.Exit.
Retention becomes a nats.Config field (default WorkQueuePolicy) so a future
reaction stream can fan out under LimitsPolicy instead of the hardcoded
competing-consumer policy that starves a second consumer on overlapping
subjects.
Adds POST /webhooks/capture/{source}: an unauthenticated endpoint (mounted
alongside /healthz, bypassing dashboard token auth by design) that persists
any inbound webhook archie has no binding for yet, so it can be inspected
and mapped later (t2db.2/t2db.3).
docs/prds/event-capture-storage.md decides SQLite over NATS JetStream for
storage: NATS is not unconditionally available until 7d5u.2 ships embedded
support, and capture's single-writer/single-reader access pattern doesn't
need JetStream's fan-out/replay machinery anyway.
internal/store/captures.go adds a captured_events table with prune-on-write
(retention age + row-count cap, both configurable). The HTTP handler applies
webhookguard redaction before persisting, and enforces two more independent
disk-bound mechanics before that: a per-request body-size cap and a
per-remote-address rate limit. New [capture] config section with documented
defaults; all fields optional, capture is on by default.
Adversarial review (manual, after the usual fresh-subagent pass hit a
session API limit mid-run) caught and fixed three real defects before this
landed, each with a regression test confirmed to fail against the pre-fix
code:
- the body-size cap silently no-op'd when CaptureMaxBodyBytes was left
unset, the one bound that limits a single payload's size rather than row
count or write rate
- the rate limiter was keyed by the request's own source path segment --
unregistered and attacker-chosen by design for this endpoint, violating
webhookguard.RateLimiter's documented bounded-key-space assumption and
letting a sender bypass the limit by rotating the segment; rekeyed to
remote address
- the retention prune compared time.RFC3339Nano strings, whose trimmed
variable-width fractional seconds break lexicographic ordering within the
same second; switched to fixed-width RFC3339
Closes archie-core-t2db.1.
config.docker.toml launches MCP servers with `npx -y <pkg>@<version>`,
and the container has no persistent npm cache, so every daemon restart
re-resolved and re-downloaded the package from the registry -- slow,
and it turns a registry outage or a yanked version into a daemon
outage. It directly amplified the crash loop observed on 2026-07-30,
since each restart repeated the whole download.
configuredMCPProvider now points npx-launched servers (matched by base
name, so an absolute-path npx still qualifies) at NPM_CONFIG_CACHE
under the daemon's own persistent work_dir, plus
NPM_CONFIG_PREFER_OFFLINE=true so a warm cache skips the registry
round trip entirely. Does not cover a package's own separate
downloads (e.g. Playwright/Puppeteer fetching a browser binary via
its postinstall script).
Closes archie-core-1786637490923-43-d4bb914b.
The identical, arguably worse instance in
internal/app/agentworker/mcp_providers.go (every per-task ephemeral
container, not just daemon restarts) needs a persistent Docker volume
that doesn't exist yet, so it's filed separately as
archie-core-qk8m rather than folded into this scoped fix.
CLI-based, env-configured, cache-less engines resolving SecretRef values
through the existing Yaegi plugin surface. Fresh resolve per call, so
values updated in the backend are picked up without daemon restart —
closing the gap that motivated the remote secret-input epic (#504),
whose value-entry web form duplicated the Bitwarden app for bws.
- examples/secret-engines/: sops (SOPS_FILE), vault (VAULT_ADDR/TOKEN,
key "path/field" or VAULT_SECRET_PATH prefix), age (AGE_FILE/
AGE_IDENTITY, key=value lines). //go:build ignore keeps them out of
the Go toolchain; yaegi treats the tag as a comment and loads them.
- enginehost: host-side Run helper registered as a Yaegi symbol. yaegi's
interpreted os/exec cannot pass env to child processes, so plugins
shell out through host code; bounded by a 30s timeout like the bws
engine.
- secret.LoadDir seeds the interpreter with the host environment: yaegi's
interpreted os package keeps a private env otherwise, and plugin
os.Getenv would silently see nothing.
- cmd/archied wires enginehost + unrestricted symbols into the engine
plugin loader, matching skillscript's interpretation posture.
- tests load the real shipped plugin files with fake CLIs on PATH;
added CONTEXT.md glossary for the remote secret-input vocabulary.
show_tool_calls gated Telegram's tool-activity narration only; the
web dashboard's SSE chat stream narrated every completed tool call
unconditionally with no way to turn it off, leaking tool names and
parameter shapes to anyone with dashboard access even when an
operator had already opted out for Telegram.
Moved ShowToolCalls from TelegramConfig to ChatConfig so it is one
setting for every channel, and wired the webui stream to it:
chatStreamSink now carries a showToolCalls flag (snapshotted once per
turn, mirroring how Telegram's liveReply snapshots it) and ToolCall
no-ops when it's off.
BREAKING: an existing [chat.telegram] show_tool_calls = true in
config.toml now parses as an unrecognized key under [chat.telegram]
and is silently ignored (TOML decoding is non-strict) rather than
erroring -- move it to the top-level [chat] table.
Chat returned only the final answer, so a wrong tool choice, a
hallucinated action, or a footgun like `systemctl stop` against
`restart` was invisible from the conversation. Only text deltas were
threaded through the turn, and tool calls executed inside Generate with
nowhere to report.
Replace the delta callback with gateway.TurnStream, which carries text
fragments and completed tool calls on one ordered path from the model
loop to the channel. The sequence is the point: a tool line that
overtakes the sentence it interrupted describes a turn that never
happened. Text-only callers use gateway.DeltaFunc and are unaffected.
A tool is reported on its result, never on its call -- the call part
carries no outcome yet, so reporting both would list every tool twice,
once with nothing to say.
Telegram renders it behind [chat.telegram] show_tool_calls, off by
default: this is debugging visibility, and it narrates every internal
step to provide it. The dashboard always shows it, being the operator's
own observability surface; its tool lines are live-only and are not
persisted with the conversation.
The Telegram live reply changes mechanism, and with it what a frame
MEANS. It used to stream through sendMessageDraft: drafts are ephemeral
30-second previews, so a tool call longer than that outlived the
preview, and successive frames were observed concatenated with their
cursors intact. It now sends one real message and edits it. An update
used to be "the next preview of the answer"; it is now "the whole
answer as it currently stands, replacing what is there". The cursor is
appended at render time and never stored, so exactly one exists and
always at the end, and a frame too long for one message is cut to its
tail rather than rejected outright.
A stopped or failed turn now edits the cursor away instead of leaving
it blinking on a message nothing will finish. That cleanup runs on a
live context because turnCtx is already dead, and it runs on both
failure branches: ai-sdk raises an aborted stream as ErrAborted with
ctx.Err() formatted in rather than wrapped, so errors.Is cannot be
trusted to tell a /stop from a fault.
Closes#467
Task and event persistence now uses a dedicated SQLite sidecar derived from the configured state prefix, matching the independently persisted conversation store. The configured path now means a state-path prefix rather than a database file so fresh SQLite state never attempts to interpret the previous storage format.
Remove the retired implementation and dependency completely, strengthen SQLite migrations and concurrency behavior, restore backend-neutral coverage, and align configuration, operational guidance, UI labels, and deployment documentation with the new storage and host-daemon topology.
internal/infrastructure/configuration/setup.Run drives operator
identity, LLM provider, models, forge, and Telegram chat channel
prompts and returns TOML edits -- not a config.Config. Building a
parallel in-memory representation to validate against was rejected
during design review: it's a second source of truth about the exact
thing (text vs. struct disagreeing) this whole feature exists to stop.
Tests instead render the edits against the real embedded template and
load the result through a real configuration.Loader, proving the
actual defaults/decode/validate path accepts it.
Two boot-time landmines in config.example.toml surfaced by that
approach, both fixed here:
- [providers.openai] ships active with api_key pointed at the bws
engine. cmd/archied/provider_secrets.go resolves every entry in
cfg.Providers at startup, not just the ones [models] references, and
refuses to start the daemon if any fails -- so picking any other
provider (or picking openai but skipping the key) left a config that
wouldn't boot on a machine without the bws CLI installed (the common
case). stepProvider now neutralises providers.openai.api_key to the
zero-value secret ref whenever openai wasn't the actual choice.
- [[repos]] shipped active with a placeholder owner/name that
validates cleanly (owner/name are only checked non-empty) and would
have archied silently polling a repo that doesn't exist. Commented
out in the template with an explanation; setup does not write this
section since tomlwrite cannot target array-of-tables.
TestExampleConfigLoads updated to expect zero repos.
Also: forge and provider secret env-var names are each defined in one
table (forgeToken, cloudProviders) read by both the config edit and
the secret sink, so the two can't name a secret differently -- the
exact bug class this feature exists to close for the installer, tested
directly (TestRun_ForgeTokenKeyMatchesStoredSecretKey compares the
decoded key against the key the secret was actually stored under, not
a hardcoded literal on both sides). Telegram writes chat.telegram.
token_env, not a secret.SecretRef -- TelegramConfig has no such field.
Telegram also requires at least one allowed_user_id: AllowedUserIDs is
deny-by-default, so a bot token with none configured answers nobody,
which reads as broken rather than as the safe default it is.
SecretSink is a caller-supplied sink (Put buffers, Commit persists),
not a return value: Run never hands back plaintext secrets, and the
caller is expected to call Commit only after proving the generated
config loads -- so a validation failure never leaves a secret written
without the config that references it.
Refs archie-core-cbk
Applies the four-reviewer pass over the forge/dashboard-actions work.
The through-line is that several capabilities were wired but inert, and
one of them could cost real work.
DUPLICATE WORK. "Closes #N" was dropped from PR bodies in favour of
LinkBranch, which is sidebar linkage on Gitea and a no-op stub on
GitHub -- neither closes anything, and nothing else did either. So a
merged task left its issue open, labelled and assigned; the next poll
re-enqueued it, and once the dashboard's Clear removed the task row that
became a second implementation and a second PR for work already merged.
reconcilePRs now closes the issue on merge, and a rejected task closes
its issue too. A PR closed without merging deliberately leaves the issue
open: that usually means "try again".
MAX RETRIES. Configured, defaulted and reported by /api/config, and
enforced nowhere -- markDead was unreachable and EffectiveMaxRetries had
no callers, so the dashboard's retry incremented retry_count forever.
The retry action now retires a task to dead at the cap.
DEAD ON ARRIVAL. Tasks() never selected plan or source, so the
"Decision required" panel could not render and chat tasks -- whose issue
number is synthetic -- linked to forge issues that do not exist.
COSMETIC FAILURE, FATAL RESULT. OpenPR failed the whole stage when
LinkBranch errored, so the pull request was never opened and the task
parked; a retry re-links the same branch, which Gitea answers with a
conflict, so every retry parked again. It is best-effort now.
INVISIBLE ACTIONS. Approve/retry/reject wrote store rows and nothing
else, so the timeline and activity stream never showed that a human had
intervened. They emit events now.
Also: store write failures answer 500 rather than 409; the dashboard
surfaces the server's actual reason instead of "409 Conflict"; action
buttons stop propagation (acting on a task no longer also expands it),
disable in flight, keep per-task errors instead of one shared string,
and confirm before rejecting.
Removed: the maybeRetryParked shim and markDead with their now-vacuous
tests, WaitingTasks (no consumers; its tests were rewritten to assert
the same routing through TaskByID), write-only Daemon.Runtime, and
dir/dirs response bloat from /api/skills. skillSource checked workDir
before sharedDir, mislabelling every shared skill when the two are equal.
DefaultRoots now dedupes roots differing only by a trailing separator.
The air build rule claiming to keep ui/dist current never ran -- it
required one file to satisfy include_dir and include_file together, and
ui/ is excluded from the walk. Deleted with the comment that promised it.
task test:ui joins task check and CI; it ran nowhere before, so the
dashboard suite could rot silently.
Docs corrected where they described behaviour that no longer exists:
parks no longer comment on the issue, state labels no longer drive
retries, and forgerpc's doc claimed four proxied methods while listing
three of five.
Test fixtures used 'sam', 'samcharles93', and 'sam/<repo>' as owner
names. These are personal and machine-specific — the repo should use
generic placeholders like 'acme' (matching the existing convention in
worktree_test.go).
config.example.toml: samcharles93 -> your-github-username
Test owner: sam -> acme
Repo refs: sam/archie-core -> acme/archie-core, etc.
Example skills: sam -> acme
/stop could not reach a running turn, and the reason was worse than a
missing cancel. go-telegram/bot runs one update worker that calls
handlers synchronously (defaultWorkers = 1, never overridden), and the
chat turn ran on it. While a turn streamed -- or sat in a shell tool --
the worker was blocked and /stop sat unread in the update channel until
the turn it was meant to kill had already finished.
Add gateway.Turns: a per-session serial lane that turns are submitted to
and that the event loop never waits on. The worker is free again in
microseconds, so /stop is serviced immediately; ordering comes from the
lane being serial rather than from the loop being blocked.
- Bare /stop cancels the running turn and clears the backlog behind it,
with no approval step. Approval guards destructive commands, and
gating the brake behind a confirmation defeats it. /stop <name> keeps
the existing process-termination behaviour and its approval.
- Cancellation propagates through the turn context, so it reaches the
model stream and any tool underneath -- a shell command built with
exec.CommandContext dies with the turn.
- Queues are unbounded and never shed. A dropped message is a silent
failure that surfaces later as an agent that ignored something.
- Turns are recovered per-lane: the channel's panic middleware wraps the
event-loop goroutine, which no longer runs the turn.
Also raise the chat step budget. It was hardcoded at 8 -- roughly one
read, one grep and two edits -- so turns hit the cap and stopped
mid-task with no indication they had been cut off. It is now
chat.max_steps, defaulting to 100. Interruption is /stop's job.
The chat agent had two tools -- memory_edit and skill_activate -- so it
could not read a file, search a tree or run a command. ai-sdk's runtime
already drives a tool loop for it (MaxSteps 8, toolExecute wired as the
executor), so the loop was never the missing piece: the registry was empty
of anything useful.
Lifted from tau internal/agent/tools at commit f5289ea, per the recorded
copy-and-mutate strategy: every file carries a provenance header naming its
tau source path, the commit, and the mutations made, so refreshing from
upstream is a diff rather than archaeology.
internal/tools/builtin/ the lifted tools, ~3.1k lines
internal/tools/provider/builtin/ the adapter to tools.ToolEntry
Six tools: read, write, edit, shell, grep, find. Tau's own tests came with
them and pass unmodified, which is the evidence that the lift is faithful.
Deliberately not lifted:
- agent.go (1214 lines): subagent spawning, and the only real coupling in
the package -- 10 tau packages including chat, sessions, store and
registry. Nothing the chat agent needs to do work.
- skill.go: depends on tau's skills package; archie has internal/skill and
already exposes skill_activate.
- docs.go: serves tau's own embedded documentation via tau/docs.
Not lifting the coordinator either. It is tau's agent loop (6.4k lines over
ten packages) and archie already has one in the ai-sdk runtime; porting it
would leave two loops competing. Its cancellation and tool events are worth
having, but they are a few hundred lines built against the existing path
rather than six thousand lifted onto a second one.
The adapter bridges two shape differences: archie hands handlers decoded
arguments where tau's executor takes raw JSON, and tau reports tool-level
failure in Result.IsError rather than as an error -- surfaced as a real
error so the runtime and the guardrail engine both see it. Tools are
classified idempotent or mutating for guardrails and budgets; that is
categorisation, not a jail.
chat.workspace roots the tools and defaults to empty, which registers none
of them. They read, write and execute, so the directory has to be a
deliberate choice rather than whatever the daemon started in -- in the
container that is /.
Verified: go build, go vet, gofmt clean; full suite passes.
Three defects found from a real Telegram transcript.
1. The operator's name was compiled into the prompt.
internal/gateway/persona.go and prompt.go both hardcoded "Sam's" into the
default persona and the fallback prompt. The persona renders first into
every system prompt, so a personal name was sent to the model provider on
every turn, and any other deployment got an assistant that believed it
worked for someone else. There was nowhere to put it: Repo.Owner is a repo
owner and config.Config had no operator concept.
ChatConfig.Operator now carries it, rendered in the prompt's <env> block as
runtime metadata and omitted entirely when unset. config.example.toml
documents that setting it sends the name to the model provider.
2. /new claimed to clear history that it had not cleared.
The chat responder keyed history by channel id through a local sessionKey
helper, while the session commands operate on tracked session ids.
sessionTracker.resolve mints deterministic ids so the two agree -- until
/new, /branch or /resume assigns a generated id, after which they address
different histories permanently. /new reported "Conversation history has
been cleared" while the next turn still saw everything, and /undo,
/compress, /resume and /sessions acted on a namespace the conversation
never read.
gateway.Router.ResolveSessionKey already existed for this, and
session_commands.go:104 said outright that main.go should call it instead of
its inline helper. It never got wired. The responder now resolves through
the router and the dead helper is deleted so it cannot come back.
buildTelegramRouter constructs the router before the responder so the
dependency is visible rather than captured by reference.
3. Prompt rules for two behaviours the transcript showed.
The agent offered to accept shell access it has no tool to use, which is
both unhonourable and a request for the user's secrets. It also opened
mid-conversation replies with a greeting, and described its own manner
("no fluff") in stock phrasing. Added core rule 7 and three communication
rules covering greetings, self-description and register.
Note the memory claim in that transcript was NOT a fabrication: the
memory_edit tool's own description names MEMORY.md and USER.md, and tool
descriptions are part of the prompt.
Also added chat-turn logging (session id, channel, thread, history depth).
Chat turns were entirely unlogged, so a report of lost context had no
evidence to check: 76 log lines covered two days and none of them was a
conversation. A separate report of mid-conversation amnesia remains
undiagnosed -- single process, 43h uptime, compression untriggered at three
messages -- and this is what makes it diagnosable next time.
Regression tests: prompt renders/omits the operator and carries no personal
name; the responder resolves via ResolveSessionKey and defines no local
session key helper.
Verified: go build, go vet, gofmt clean; full suite passes.
Follow-up to c924281, which introduced two defects of its own.
1. indexing.Config was a private config island: index paths existed nowhere
in internal/config or config.example.toml, so they were undiscoverable
and unsettable by anyone reading the real configuration.
Now [indexing] in config.toml, defined in its own internal/config/
indexer.go rather than appended to the 21-struct config.go, and its
defaulting lives with it as IndexingConfig.withDefaults so finalize
delegates instead of growing. Paths default off work_dir, keeping the
two archied instances on one host from sharing an index. Documented in
config.example.toml.
internal/indexing keeps its own plain Config and does NOT import
internal/config, so it stays a straight diff against its tau original.
2. A failed index build was invisible. refreshAsync wrote the reason to a
database column with no reader and degraded to unindexed search in
silence — the failure mode that costs hours precisely because nothing
appears to be wrong. Every give-up path now logs, and a successful
build logs once.
The logger is resolved at the point of use, not in the constructor: a
Manager can be built as a struct literal (the ported tests do), and
constructor-only defaulting panicked on the build goroutine, where a panic
takes the daemon with it. Regression test covers the zero and nil Manager.
Adds internal/secret: a pluggable Engine interface (Name/Version/Resolve)
resolved through a Registry, so secrets backends (sops, Bitwarden Secrets
Manager, Vault/OpenBao) aren't hard-coded into the binary. Only the "env"
engine is compiled in; custom engines are Yaegi-interpreted .go plugins
loaded via Registry.LoadDir, mirroring internal/plugin's conventions
(secretextract provides the generated, nil-guarded symbol table for the
Yaegi/Go interface boundary).
config.Forge.TokenEnv is replaced end-to-end with config.Forge.Token
(secret.SecretRef{Engine, Key}) -- finalize() defaulting, forge token
resolution in cmd/archied/main.go, and per-identity validation that now
requires an explicit token instead of silently resolving to "". No
backwards-compat shim.
Provider.APIKeyEnv, NATSConfig.TokenEnv, and TelegramConfig.TokenEnv are
deliberately not migrated here -- they're forwarded by name to
ai-sdk/runtime and subprocess workers rather than resolved centrally,
which needs its own eager-resolve-then-inject design pass
(archie-core-xec.4). The three shipped engine plugins (sops, bws,
vault/openbao) are tracked separately as archie-core-xec.1-3.
Adversarially reviewed by a fresh agent with no implementation context;
fixed the two findings that survived (missing per-identity token
validation, underdocumented nil-wrapper Resolve edge case).
Add [containers].volume_ttl (default 72h). Persistent-storage repos get a
per-repo bare Git cache (--reference-if-able --dissociate) so task
worktrees clone fast without repeated full network clones, while staying
fully independent of the cache's lifecycle — cache expiry can never
corrupt an active or parked task's worktree. Docker repo volumes are
labeled as Archie-owned and swept non-forced (a volume still attached to a
running container survives and retries next cycle); the repo cache
directory is swept by mtime. Both cleanups run at daemon startup and each
poll cycle.
Filed archie-core-756 (repo-wide golangci-lint baseline) and archie-core-340
(intermittent JetStream TempDir teardown flake) as separate follow-ups
rather than scope-creeping them into this change.
Add token_env to [nats] config, resolved and validated at daemon startup
(fails early if configured but the named env var is unset). Container
workers get NATS_URL/NATS_TOKEN; archie-agent honors both while CLI-supplied
URL still takes precedence. Anonymous auth remains the default for
single-node deployments.
Unblocks archie-core-5zh (promote archie-agent to per-task process).
Add config.Repo.AllowConcurrent (allow_concurrent toml key). Repos opted in
skip the daemon's default FIFO same-repo task serialization while still
respecting [containers].max_concurrency globally. Worktrees are already keyed
per issue number, so opted-in repos don't collide on disk.
Daemon previously processed one task at a time. Add bounded concurrent
dispatch for SQLite and NATS agent modes, reusing [containers].max_concurrency
(0 = unlimited). Same-repository tasks stay FIFO-serialized to avoid worktree
conflicts; different repositories now run concurrently.
Unblocks archie-core-l1q (NATS transport) and archie-core-g1l
(per-repo allow_concurrent override).
The first feature run parked exactly as designed but for an avoidable
reason: todo is a templ project and the agent edited generated
_templ.go files, leaving literal template placeholders in rendered
HTML. Systemic fixes, all environmental:
- repos gain protect = ["_templ.go"]: suffixes agents can never write
(agentloop.ProtectPaths), forcing edits to the sources; the gate's
new leading ["templ","generate","./components/"] command
regenerates. AgentStage.ProtectPaths now composes with the repo list
and takes TaskContext like every other stage hook.
- archied -requeue <task-id>: the retry path for parked/waiting tasks
(keeps the task's workflow; Requeue with empty workflow preserves it,
clears the park reason).
Retry of the parked feature task drove the full loop home: PRD-guided
plan -> builder edited components.templ (generated file write-blocked),
gate regenerated + tested -> PR #8 with 109 lines of tests.