Commit Graph
139 Commits
Author SHA1 Message Date
sam c35d9b6fdd feat(messaging): Messaging Service operator surface, chat report rendering, and the diff-cap config fix (#899)
* feat(workflow): post line-anchored review comments with suggested fixes

Bead archie-core-q9au (docs/prds/inline-review.md). A review's findings reached
the operator only as a flat list in the PR body, so the author had to map every
finding back to its line by hand and apply mechanical fixes by hand. Now the
line-anchored findings are also posted as inline review comments, carrying a
one-click suggestion when the reviewer could state the fix.

Domain (internal/domain/workflow):
- ReviewFinding gains Suggestion; Validate rejects one on anything but a
  confirmed finding, so the agent's record_finding call is refused with
  something it can act on. Validate splits into validateAnchor /
  validateClassification / validateDisposition, which also brings it back under
  the cyclop limit the gate enforces.
- StagePostReviewComments runs after StageOpenPR in the implement workflow. It
  posts only line-anchored findings (Line > 0); a whole-file finding has no line
  to attach to and stays in the body list, which remains the complete record.
  It is best-effort: the PR is open and the body already lists every finding, so
  a failed comment is logged, never a park -- the same reasoning as OpenPR's
  existing best-effort LinkBranch call.
- The suggestion fence is gated on Verdict == confirmed where the fence is
  written, because it is a one-click apply: a wrong suggestion breaks the code
  and the author has to notice.

Forge (internal/forge): ReviewCommentWriter + GitHub/Gitea implementations,
type-asserted like PullRequestReader so a forge that cannot is skipped rather
than faked (NoopForge is pinned as not satisfying it). GitHub loops one
CreateComment per comment and joins the failures, so one refused anchor cannot
drop the rest; Gitea carries them on a single COMMENT-state review.

Two deliberate departures from the PRD's sketch, both recorded in it:
- The interface is batch (CreateReviewComments), not one call per comment:
  Gitea's only inline shape is one submitted review, and the set then crosses
  NATS once instead of once per finding.
- The head SHA is resolved inside each forge implementation rather than by the
  caller. The stage runs in archie-agent, which holds no forge credentials and
  has no PullRequestReader; the daemon-side forgerpc handler is the only place
  that does.

Transport (internal/forgerpc): workflow.Forger gains the method, so the agent
reaches the daemon over the new create_review_comments subject; the server
type-asserts the writer and answers an incapable forge with an error the stage
logs. This is the one proxied method whose absence on an image skew does not
park a task.

The reviewer prompt and record_finding schema ask for a suggestion on a
confirmed, mechanically fixable defect only.

Regenerating the wfextract symbol table -- the documented way a new exported
workflow symbol reaches interpreted stage plugins -- also picked up symbols that
were missing from the committed table (the Status* and Source* constants,
ReviewCheck, Store, Task, and the _Store wrapper). Nil-guards were re-applied to
every wrapper afterwards, as that file's header requires.

Evidence: task check exit 0 (incl. 229 UI tests). Three mutations of the new
stage logic each fail a test that names the behaviour -- dropping the
confirmed-only gate, dropping the Line > 0 filter, and letting a posting failure
escape -- so the tests discriminate rather than passing vacuously.

* fix(build): drop the Dockerfile pin assertion

The guard fetched golangci-lint's go.mod at image build time to fail the build
when GOLANGCI_LINT_VERSION and GOFUMPT_VERSION disagreed. It only ever tripped
on self-inflicted pin drift, but it added a network fetch and a new failure
mode to the release build, which is not where that belongs.

The pins stay correct (golangci-lint 2.13.2, whose vendored gofumpt v0.11.0
equals GOFUMPT_VERSION) and the comment above them still states the invariant,
so the reasoning survives without the build-time machinery. The pin change
itself reaches the agent image through the normal release path: CI rebuilds
and pushes archie-agent on an archie/vX.Y.Z release, and a gateway-only
release deliberately does not.

* feat(workflow): add the remediate workflow for PR review remediation

Implements step 5 of docs/prds/pr-review-remediation.md: a remediate
workflow that resumes the task's existing worktree onto its PR branch
(worktree.Manager.Resume, not Prepare's refresh), runs a builder agent
against the review's actionable comments, commits and pushes to the
same branch, and replies via the forge's ReplyToReview. Task.RetryCount
doubles as the remediation round counter, capped by the repo's
max_retries; the round-cap stage parks the task and posts one comment
when it's hit.

Task gains a ReviewPayload field (JSON-encoded review unit) with full
store/proto/staterpc plumbing. workflow.Forger gains Comment and
ReplyToReview, proxied to archie-agent via a new forgerpc RPC. Resume
and its Dir lookup are added as an unexported optional Trees capability
rather than widening the exported interface, which is already at
golangci-lint's interfacebloat cap and is projected into wfextract's
Yaegi symbol table for repository-authored stage plugins.

The daemon-side reaction consumer that dispatches into this workflow
(subscribing ARCHIE_REACTIONS, grouping comments by review, enforcing
per-task serialization) is tracked separately in archie-core-8li9.4.

* feat(messaging): extract standalone messaging service binary and process composition

* feat(messaging): adapt channel frontends to consume Gateway ChatContract

* feat(messaging): remove in-process channels and Router from daemon bootstrap

Telegram, email and webhook now run only in the Messaging Service, which
reaches this process through the Gateway's ChatContract. The daemon builds
no channel and no gateway.Router, so the wiring that existed only to serve
them goes with it: setupGateways, setupEmailGateway, setupWebhookGateway,
setupTelegramGateway, buildTelegramRouter, makeTelegramReload and the
release-announcement path helper.

The seams those paths owned move to where they can still be honoured:

- Channel ValidateConfig runs in archiemessaging.compose, which now returns
  an error, so a configured-but-invalid front-end fails the service rather
  than being dropped silently.
- The inbound rate limiter moves to the Gateway process, which owns the only
  Router and therefore the only place a per-sender budget applies to every
  channel's turns.
- The daemon's gateway readiness probe becomes a ContractProbe against the
  Gateway. It cannot observe a channel lifecycle it no longer runs, and a
  channel-count probe would have reported every daemon degraded.
- telegramSetup is now chatSetup (in chat_setup.go), trimmed to the fields a
  chat turn runner actually reads.

Telegram operator features that were wired only in the daemon (/update,
release announcements, version reporting, SIGHUP reload) are not yet wired
in the Messaging Service: archie-core-8cda.6.6. Channel lifecycle reporting:
archie-core-8cda.6.8. The Gateway process builds no worktree manager, so
review_pr is unadvertised on every channel: archie-core-8cda.6.7.

* feat(messaging): re-wire the Telegram operator surface in the Messaging Service

setupTelegramGateway satisfied eight telegram.Gateway seams from the daemon's
own process. Each is resolved by where the fact it needs is owned, not by
where it used to be built; the decision table is in
docs/architecture/migration-decisions.md.

Wired here: /update (built from this service's own [chat.telegram] check and
install commands), the update-report relay (same sha256(bot_user)[:8] path the
daemon wrote, so a pending report is found rather than written afresh),
show_tool_calls, and /restart's config reload, which re-resolves the token and
allowlist from this service's own file and refuses a reload that resolves to
an empty token.

/version is sourced from the Gateway: ChatSnapshot gains a Version field
carrying the Gateway's rendered build block, read per call and bounded by the
dependency timeout. cmd/archie-messaging deliberately takes no version
ldflags, so it cannot report a partial upgrade as a matched one.

RunningVersions and ReleaseAnnouncements stay nil. Both need a component's
real running version, and the only versions this process can obtain are a
different binary's stamps -- reporting them would manufacture exactly the
false success the update check exists to catch, and the announcer without
parseable versions is a no-op that reads as configured. archie-core-8cda.6.9
carries the contract that unblocks them. Dangerous stays nil: no composition
has ever supplied a command authority.

Enrich is dropped from the update service: componentInstallTypeEnricher read
[nats], which this service must not decode, so the check command's own
reported install type stands.

* fix(workflow): guard inline review comments against head drift

Two findings from the review of q9au.

A comment's line numbers describe exactly one revision of the branch, but the
forge resolved the pull request's head at posting time -- after the review had
already read the worktree. A push in that window (a collaborator, a retry) meant
the findings were anchored to whatever now occupied those lines, mislabelling
instead of failing. The revision the review measured now travels with the
comment set: OpenPR records it from the same worktree read it already performs
(TaskContext.ReviewedHeadSHA), StagePostReviewComments threads it through
Forger to forgerpc, and the forge refuses the whole set once the head it reads
has moved off it -- which the best-effort stage logs, the PR body remaining the
record. An unmeasured revision posts unverified rather than dropping every
finding, so a worktree that cannot report one degrades instead of regressing.

Suggestion was documented single-line only, because GitHub applies a suggestion
fence up to its first line and a multi-line replacement would silently corrupt
the file. Only the verdict was validated, so a multi-line suggestion reached the
renderer and posted as that truncated fence. Validate now rejects an embedded
newline, which makes record_finding refuse it with feedback the model can act on
instead of the reviewer discovering the corruption later.

* fix(gateway): render chat reports as lines, not one run-on paragraph

/status and /tasks are line-oriented reports, but their newlines were plain
"\n" -- CommonMark's soft wrap. The Telegram rich-block renderer space-joins
soft-wrapped lines, by design: a Telegram paragraph renders an embedded
newline as nothing at all, so a space is the only separator that survives.
The reports therefore arrived as one line each: "Runtime Provider: DeepSeek
Model: deepseek/deepseek-flash", and every task's title, state and park
reason jammed together.

The renderer now implements CommonMark's hard line break (two trailing
spaces, or a trailing backslash): the line closes its own block and the next
one hugs it, so a stanza keeps its lines without the spacer paragraph that
separates unrelated blocks. Ordinary prose is untouched -- a model's wrapped
paragraph still joins with a space. The producers mark their report lines
with it.

Two report fixes alongside:

- A park reason is capped to its first line at 160 runes. It carries whatever
  the failing stage reported, which for a lint gate is the whole tool log;
  unbounded, one parked task pushed every other task out of the reply. The
  full reason is still on the task.
- /status loses its "Runtime" sub-header. A section header over two fields
  made a four-fact report read like a document.

* fix(config): let diff_cap_lines = 0 actually switch the cap off

Config.DiffCapLines was a plain int, so the defaults pass could not tell an
absent key from an explicit 0 and rewrote both to 400. The dashboard schema
said "0 means unlimited" and StageDiffCap treated <= 0 as unlimited, but 0
never reached the step: the documented way to switch the cap off was
unreachable through config.

Same root cause, worse instance: configForIdentity assigned
root.DiffCapLines = identity.DiffCapLines unconditionally, so an identity
that simply omitted the key overwrote the shared cap with the zero value and
silently disabled the safety rail for that identity.

DiffCapLines is now *int on both Config and IdentityConfig -- nil is absent,
0 is no cap -- read through Config.DiffCap(). Defaults fill only nil, and an
identity overrides only when it set the key. TaskConfig keeps a resolved int,
so the daemon/agent wire format is unchanged and a version-skewed agent
cannot misread a null.

The park reason changes too. It told the operator to "approve manually", but
a parked task's actions are retry, abandon and reject (internal/taskstate):
there is no approve path out of diff-cap, so it sent them looking for a
button that does not exist. It now names the two things that work.

* feat(config): make the daemon's readiness dependency timeout configurable

The daemon's gateway probe bounded its call with a hardcoded 5s constant,
while every other NewContractProbe call site (archie-ui, archie-messaging)
passes the service's configured DependencyTimeout. It was the only magic
number among them.

[health].dependency_timeout carries it, defaulting to the same 5s the
extracted services' -dependency-timeout flag defaults to, so every process
bounds a probe into a dependency the same way out of the box.

It is not added to the dashboard config schema or the reloadable set:
[health] appears in neither today, health.listen included, and a probe
assembled at boot would not observe a reload anyway.
2026-09-20 00:32:13 +10:00
sam 800352a692 feat(gateway): give /status a real daemon-health surface (#892)
* feat(gateway): give /status a real daemon-health surface

Phase 2 of archie-core-wp9s. Phase 1 split the commands apart (health vs
work) but left /status with one aggregate queue line and a static runtime
block; it could not answer the questions an operator actually asks when
something feels wrong. /status now reports, between the queue line and the
runtime block:

- Broker: the live state of this process's own NATS connection
  (eventbus/nats.Client.Connected), not a fresh dial -- a probe would
  report the probe's reachability, not the daemon's.
- Containers: the pool's own active counter against the configured cap
  (Pool.Active/Cap), not a Docker listing, which would include containers
  this pool does not own. A zero cap means unlimited, so it is not shown
  as a ratio.
- Channels: the channel manager's real lifecycle state, including a
  failed adapter's reason (the documented decision: read the existing
  manager rather than re-deriving health per adapter).
- Chat model: last-known outcome of every chat-model call this process
  made, recorded once in sendChatTurn (the single choke point) and read
  back by /status. The second documented decision: never a live probe,
  so /status stays instant and reports what happened -- or says plainly
  that nothing has been attempted yet.
- Last poll: when the most recent poll pass began, stamped on both poll
  paths (Run's single-identity loop and runIdentities' per-identity
  goroutines). A pass that hangs leaves the stamp stale, which is the
  signal; stamping completion would leave a wedged poller looking
  healthy.

Truthful or absent: a section this process has no source for is omitted
rather than rendered as a zero. The standalone Gateway owns no poll loop,
pool or channel manager, so its /status reports broker and chat-model
health only; aggregating across processes would need a new daemon RPC and
is deliberately not built. Version stays out (/version owns it).

/status and /tasks descriptions now describe what each command actually
does on both published surfaces (Telegram menu/help and the shared
specs), pinned by a test that also fails if the two surfaces drift.

Sources are read through functions, not captured values: occupancy and
channel state move, and the daemon is built after the gateways, so a
source captured at construction would freeze /status on boot state.

Left in place deliberately: the Runtime (provider/model) block, which
Phase 1 flagged and which duplicates /model and /whoami. Removing it is a
product call on a command Sam uses daily, not a mechanical follow-up.

* docs(prds): record the two open /status health decisions as settled

Both were flagged in PR #892 as needing the maintainer's call; both are now
answered, and the PRD is the place a follow-up would look:

- The Runtime (provider/model) block stays. It is the one line in /status the
  ticket's own rule says should not be there, but /status is read daily to see
  the active model at a glance, so the duplication is deliberate. Recorded as a
  settled product decision so it is not removed later as an oversight.
- Health stays per-process. Pooling the daemon's and the Gateway's chat-model
  outcomes would need a new daemon<->gateway health RPC for one status line, and
  a per-process line is true where an aggregate would be a claim about a process
  the answering one cannot see.

/agents' published description is still left to archie-core-mxls, unchanged.
2026-09-20 00:30:23 +10:00
sam e0b6aa6105 feat(workflow): post line-anchored review comments with suggested fixes (#891)
* feat(workflow): post line-anchored review comments with suggested fixes

Bead archie-core-q9au (docs/prds/inline-review.md). A review's findings reached
the operator only as a flat list in the PR body, so the author had to map every
finding back to its line by hand and apply mechanical fixes by hand. Now the
line-anchored findings are also posted as inline review comments, carrying a
one-click suggestion when the reviewer could state the fix.

Domain (internal/domain/workflow):
- ReviewFinding gains Suggestion; Validate rejects one on anything but a
  confirmed finding, so the agent's record_finding call is refused with
  something it can act on. Validate splits into validateAnchor /
  validateClassification / validateDisposition, which also brings it back under
  the cyclop limit the gate enforces.
- StagePostReviewComments runs after StageOpenPR in the implement workflow. It
  posts only line-anchored findings (Line > 0); a whole-file finding has no line
  to attach to and stays in the body list, which remains the complete record.
  It is best-effort: the PR is open and the body already lists every finding, so
  a failed comment is logged, never a park -- the same reasoning as OpenPR's
  existing best-effort LinkBranch call.
- The suggestion fence is gated on Verdict == confirmed where the fence is
  written, because it is a one-click apply: a wrong suggestion breaks the code
  and the author has to notice.

Forge (internal/forge): ReviewCommentWriter + GitHub/Gitea implementations,
type-asserted like PullRequestReader so a forge that cannot is skipped rather
than faked (NoopForge is pinned as not satisfying it). GitHub loops one
CreateComment per comment and joins the failures, so one refused anchor cannot
drop the rest; Gitea carries them on a single COMMENT-state review.

Two deliberate departures from the PRD's sketch, both recorded in it:
- The interface is batch (CreateReviewComments), not one call per comment:
  Gitea's only inline shape is one submitted review, and the set then crosses
  NATS once instead of once per finding.
- The head SHA is resolved inside each forge implementation rather than by the
  caller. The stage runs in archie-agent, which holds no forge credentials and
  has no PullRequestReader; the daemon-side forgerpc handler is the only place
  that does.

Transport (internal/forgerpc): workflow.Forger gains the method, so the agent
reaches the daemon over the new create_review_comments subject; the server
type-asserts the writer and answers an incapable forge with an error the stage
logs. This is the one proxied method whose absence on an image skew does not
park a task.

The reviewer prompt and record_finding schema ask for a suggestion on a
confirmed, mechanically fixable defect only.

Regenerating the wfextract symbol table -- the documented way a new exported
workflow symbol reaches interpreted stage plugins -- also picked up symbols that
were missing from the committed table (the Status* and Source* constants,
ReviewCheck, Store, Task, and the _Store wrapper). Nil-guards were re-applied to
every wrapper afterwards, as that file's header requires.

Evidence: task check exit 0 (incl. 229 UI tests). Three mutations of the new
stage logic each fail a test that names the behaviour -- dropping the
confirmed-only gate, dropping the Line > 0 filter, and letting a posting failure
escape -- so the tests discriminate rather than passing vacuously.

* fix(build): drop the Dockerfile pin assertion

The guard fetched golangci-lint's go.mod at image build time to fail the build
when GOLANGCI_LINT_VERSION and GOFUMPT_VERSION disagreed. It only ever tripped
on self-inflicted pin drift, but it added a network fetch and a new failure
mode to the release build, which is not where that belongs.

The pins stay correct (golangci-lint 2.13.2, whose vendored gofumpt v0.11.0
equals GOFUMPT_VERSION) and the comment above them still states the invariant,
so the reasoning survives without the build-time machinery. The pin change
itself reaches the agent image through the normal release path: CI rebuilds
and pushes archie-agent on an archie/vX.Y.Z release, and a gateway-only
release deliberately does not.

* feat(workflow): add the remediate workflow for PR review remediation

Implements step 5 of docs/prds/pr-review-remediation.md: a remediate
workflow that resumes the task's existing worktree onto its PR branch
(worktree.Manager.Resume, not Prepare's refresh), runs a builder agent
against the review's actionable comments, commits and pushes to the
same branch, and replies via the forge's ReplyToReview. Task.RetryCount
doubles as the remediation round counter, capped by the repo's
max_retries; the round-cap stage parks the task and posts one comment
when it's hit.

Task gains a ReviewPayload field (JSON-encoded review unit) with full
store/proto/staterpc plumbing. workflow.Forger gains Comment and
ReplyToReview, proxied to archie-agent via a new forgerpc RPC. Resume
and its Dir lookup are added as an unexported optional Trees capability
rather than widening the exported interface, which is already at
golangci-lint's interfacebloat cap and is projected into wfextract's
Yaegi symbol table for repository-authored stage plugins.

The daemon-side reaction consumer that dispatches into this workflow
(subscribing ARCHIE_REACTIONS, grouping comments by review, enforcing
per-task serialization) is tracked separately in archie-core-8li9.4.

* feat(messaging): extract standalone messaging service binary and process composition

* feat(messaging): adapt channel frontends to consume Gateway ChatContract

* feat(messaging): remove in-process channels and Router from daemon bootstrap

Telegram, email and webhook now run only in the Messaging Service, which
reaches this process through the Gateway's ChatContract. The daemon builds
no channel and no gateway.Router, so the wiring that existed only to serve
them goes with it: setupGateways, setupEmailGateway, setupWebhookGateway,
setupTelegramGateway, buildTelegramRouter, makeTelegramReload and the
release-announcement path helper.

The seams those paths owned move to where they can still be honoured:

- Channel ValidateConfig runs in archiemessaging.compose, which now returns
  an error, so a configured-but-invalid front-end fails the service rather
  than being dropped silently.
- The inbound rate limiter moves to the Gateway process, which owns the only
  Router and therefore the only place a per-sender budget applies to every
  channel's turns.
- The daemon's gateway readiness probe becomes a ContractProbe against the
  Gateway. It cannot observe a channel lifecycle it no longer runs, and a
  channel-count probe would have reported every daemon degraded.
- telegramSetup is now chatSetup (in chat_setup.go), trimmed to the fields a
  chat turn runner actually reads.

Telegram operator features that were wired only in the daemon (/update,
release announcements, version reporting, SIGHUP reload) are not yet wired
in the Messaging Service: archie-core-8cda.6.6. Channel lifecycle reporting:
archie-core-8cda.6.8. The Gateway process builds no worktree manager, so
review_pr is unadvertised on every channel: archie-core-8cda.6.7.

* feat(messaging): re-wire the Telegram operator surface in the Messaging Service

setupTelegramGateway satisfied eight telegram.Gateway seams from the daemon's
own process. Each is resolved by where the fact it needs is owned, not by
where it used to be built; the decision table is in
docs/architecture/migration-decisions.md.

Wired here: /update (built from this service's own [chat.telegram] check and
install commands), the update-report relay (same sha256(bot_user)[:8] path the
daemon wrote, so a pending report is found rather than written afresh),
show_tool_calls, and /restart's config reload, which re-resolves the token and
allowlist from this service's own file and refuses a reload that resolves to
an empty token.

/version is sourced from the Gateway: ChatSnapshot gains a Version field
carrying the Gateway's rendered build block, read per call and bounded by the
dependency timeout. cmd/archie-messaging deliberately takes no version
ldflags, so it cannot report a partial upgrade as a matched one.

RunningVersions and ReleaseAnnouncements stay nil. Both need a component's
real running version, and the only versions this process can obtain are a
different binary's stamps -- reporting them would manufacture exactly the
false success the update check exists to catch, and the announcer without
parseable versions is a no-op that reads as configured. archie-core-8cda.6.9
carries the contract that unblocks them. Dangerous stays nil: no composition
has ever supplied a command authority.

Enrich is dropped from the update service: componentInstallTypeEnricher read
[nats], which this service must not decode, so the check command's own
reported install type stands.

* fix(workflow): guard inline review comments against head drift

Two findings from the review of q9au.

A comment's line numbers describe exactly one revision of the branch, but the
forge resolved the pull request's head at posting time -- after the review had
already read the worktree. A push in that window (a collaborator, a retry) meant
the findings were anchored to whatever now occupied those lines, mislabelling
instead of failing. The revision the review measured now travels with the
comment set: OpenPR records it from the same worktree read it already performs
(TaskContext.ReviewedHeadSHA), StagePostReviewComments threads it through
Forger to forgerpc, and the forge refuses the whole set once the head it reads
has moved off it -- which the best-effort stage logs, the PR body remaining the
record. An unmeasured revision posts unverified rather than dropping every
finding, so a worktree that cannot report one degrades instead of regressing.

Suggestion was documented single-line only, because GitHub applies a suggestion
fence up to its first line and a multi-line replacement would silently corrupt
the file. Only the verdict was validated, so a multi-line suggestion reached the
renderer and posted as that truncated fence. Validate now rejects an embedded
newline, which makes record_finding refuse it with feedback the model can act on
instead of the reviewer discovering the corruption later.
2026-09-19 23:37:23 +10:00
sam 04d4d8f1d8 chore(memory): delete internal/memory (unification slice 5) (#889)
Deletes internal/memory (Manager, builtin.Provider, memory_edit tool) now
that slices 2-4 built the CRUD engine family and its per-turn read/write
seams. Also deletes internal/infrastructure/memory/builtin/provider.go (the
memory_edit adapter over the old package -- its markdown store already
relocated in slice 2), internal/tools/provider/memory (the old tool
registration, redundant with slice 4's memory_create/update/delete/list),
config.Memory.{Provider,ProviderConfig} and the validateMemory rejection
that existed only to complain about Provider, and webui.MemoryStatus/
MemoryProviderHandle plus its dashboard page -- never wired in production
(s.Memory was always nil), confirmed by /api/capabilities always reporting
"memory": false.

Rewrites ARCHITECTURE.md's "Memory engine family" section to describe the
current engine (internal/domain/memory contract, internal/infrastructure/
memory builtin engine + its builtin/ markdown store, the per-turn read seam
in gateway/turn_memory.go, the per-turn write tools in
gateway/turn_memory_tool.go) instead of the deleted Manager. Fixes two
docs/architecture/configuration.md rows that described the deleted
manager's literals and SaveConfig mechanism.
2026-09-19 13:43:16 +10:00
sam 7ee5025d3c feat(memory): four-scope contract and builtin engine CRUD (unification slice 2) (#886)
* feat(memory): four-scope contract with CRUD and revisions

Seam for memory unification slice 2 (docs/prds/memory-engine-unification.md
§1, §2, §7): internal/domain/memory is now a scope-addressed CRUD contract.

- Scope{Kind,Agent,User} with the four ratified kinds, Validate rejecting a
  component present where it is not required, and a length-prefixed Key so
  opaque channel-native ids cannot collide.
- Subject{AgentID,UserID} with Scopes()/WritableScopes(): global is never
  writable, and an absent user id removes every scope requiring one.
- Store becomes Create/Get/Query/List/Update/Forget/Revisions over
  NewRecord/RecordUpdate/Record/Revision, with ErrNotFound/ErrStaleRevision.
  Observation, Identity and Metadata are deleted.
- The content scanner moves into the family (internal/domain/memory) so every
  engine applies it in Create and Update; internal/memory re-exports the types
  until slice 5 deletes it.
- The markdown store moves to internal/infrastructure/memory/builtin, its
  correct home per organisation.md.

WIP: internal/infrastructure/memory and internal/infrastructure/sessioncurator
are converted in the lanes that follow this commit.

* feat(memory): builtin engine CRUD, retained revisions and the curator's agent-user write

Slice 2 implementation (docs/prds/memory-engine-unification.md §1, §2, §6,
§7), on top of the contract seam in the previous commit.

- internal/infrastructure/memory: per-scope markdown heads with a JSON
  provenance marker, per-scope HISTORY.md retaining superseded and deleted
  states, Update superseding before replacing the live block, ErrStaleRevision
  via RecordUpdate.Expected, idempotent Forget, and the family scanner applied
  in Create and Update. Scope directories are hex(sha256(Scope.Key())).
- internal/infrastructure/sessioncurator: the participant is derived from the
  session's user-role senders; exactly one writes agent-user scope, zero or
  several write nothing and return ActionSkipped naming the count.
- docs/architecture: the settled decision recorded in migration-decisions §5
  (superseding the 2026-09-02 marker entry) and agent-system.md pointing at it.

Draft: to be squashed and reviewed before publication.

* test(curator): conform the registry's fake memory engine to the CRUD contract

The fake still implemented the pre-CRUD Store surface (Write taking an
Observation, List/Forget keyed by a bare string), so the package's test
binary no longer compiles: fakeMemoryEngine is missing Create, Get, Update
and Revisions, and domainmemory.Observation no longer exists. That took
`go fix ./...`, `go vet ./...` and `go test ./...` down with it, which is
what `task check` runs first.

Missed by the same commit that updated internal/infrastructure/sessioncurator's
own fake to the new interface. Repaired on the fake, not by loosening the
interface: the fake is the only thing out of date.

* fix(sessioncurator): refuse to attribute a session with unidentified user messages

sessionParticipant skipped user-role messages with an empty SenderID and
counted only the identified ones, so a session holding one identified sender
plus any number of unidentified user messages yielded that sender as the
single participant. Those messages are still part of the excerpt the
extraction model reads, so facts derived from content of unknown provenance
were written into the identified sender's agent-user scope: user A's memory
can end up crediting content that may belong to someone else. It is exactly
the case the function's own doc comment claims to handle, which it only
caught when no message at all was identified.

An unidentified user message now makes the session unattributable outright,
beside any number of identified senders, and the skip action's reason names
the count instead of reporting a sender count that no longer explains the
skip. Assistant-role messages are unaffected: they carry the bot, so they
still cannot manufacture a participant.

Fixed in sessionParticipant rather than at the write, because it is the
single producer of the attribution decision: it now returns the participant
or the reason there is none, so the caller cannot print a reason that
disagrees with what it decided. The fail-closed path (nothing written, one
ActionSkipped) and the extraction flow are unchanged.

* fix(memory): carry record content through the block format intact

builtin.Store's format ends a block at a blank line and starts a section at a
"## " line, but storeContent only trimmed the content and rejected an empty
one. Content holding either delimiter was accepted, written, and answered back
to the caller in full -- then read back truncated at the delimiter by Get,
List, or a process restart, with the remainder unaddressable and no error
anywhere along the path.

Fixed at the producer, in the render/parse pair that owns the block format:
renderBlock escapes the two delimiter shapes (an empty line becomes a single
backslash, a line already starting with a backslash or "## " gains one) and
parseBlocks undoes it on the way back, so a record's content survives the
write whole. Rejecting the content instead would have made an ordinary
multi-paragraph note or a quoted heading unrepresentable; escaping every line
would have changed the on-disk shape of every record. Only the lines that need
it are escaped, so a one-paragraph record is byte-identical to what the
unfixed engine wrote and the document stays hand-editable.

Reachable from both producers: the chat tool's note and the curator's
model-extracted fact are free text, and neither is constrained to one
paragraph.

Tests: TestBuiltinEngineRenderedBlockRoundTripsContent asserts
parseBlocks(renderBlock(c)) == c for eleven content shapes, and
TestBuiltinEngineContentSurvivesTheWriteAndAReopen carries the same shapes
through Create, Get, and a second engine reading the file a restart would.

* fix(memory): compare Update's expected revision against the file, not the cache

Update read the record to compare against out of findBlock(stores.live, ...)
-- the engine's cached copy of the scope document -- and checked Expected
against that copy. This engine caches one document per scope and rewrites it
whole, so a second engine over the same scope file (archied and the gateway,
or a restart) holds a copy taken when it last read the file: its view of the
record's revision is superseded, Expected matches the revision it remembers,
ErrStaleRevision never fires, and the rewrite that follows writes that stale
copy back over the file -- discarding the other writer's update and every
record written since the copy was loaded. The field existed and could not
fire.

Reload the scope document before anything is decided from it, so the check
answers about the revision actually on disk and the rewrite starts from the
current document. That is the smaller of the two fixes the PRD leaves open
("Not determined": a reload-before-mutate or an advisory lock); no locking
scheme, no retry loop, and no change to the read paths.

Still unsolved, and deliberately: the whole-document rewrite is
last-writer-wins between the reload and the write, so two engines writing one
scope concurrently can still lose one of the two writes, and HISTORY.md is
rewritten from the cached copy the same way.

Tests: TestBuiltinEngineSecondEngineCannotReuseAStaleRevision asserts the
second engine gets ErrStaleRevision and the first writer's state survives;
TestBuiltinEngineUpdateDoesNotEraseRecordsItHasNotLoaded asserts the record
written after the second engine loaded is still there after its update.

* fix(memory): emit the scanner's warning instead of dropping it

scanContent returned nil for every result that was not ThreatBlock, so a
ThreatWarn result -- documented in the scanner as "allow the write but emit a
warning", and what every sensitive-data match produces (API keys, passwords,
tokens, private keys) -- was stored with no audit signal anywhere. The
comment justified the silence by pointing at Bind, but Bind's argument is
about Registrar.Events: internal/events declares no memory kind and the
composition binds no memory event sink, so an invented kind would look wired
and not be. A scanner warning is a diagnostic rather than an event, and there
was nothing else to carry it: the registrar had no logger.

Bind now takes Registrar.Log, optional and mirroring the curator family's
Registrar for the same reason, scanContent emits the warn-level result
through it before the write is allowed, and bootstrap passes the daemon's
logger so the line reaches archied's configured sinks instead of only the
process default. The write is still allowed: only the silence changed.

Tests: TestBuiltinEngineScanWarnsWithoutRefusingTheWrite runs Create and
Update over warn, clean and block content, asserting the write lands either
way, that a warn-level scan result logs exactly one warning carrying the
scanner's own message, that clean content logs none, and that a refused
block-level match logs none.

* test(memory): carry every content shape through Update and Revisions

Update replaces the block it read out of the document by that text, and
retains the superseded state in HISTORY.md, so an escaped block has to be
found, superseded and retained exactly like an unescaped one. Covered here
rather than left to the Create path: without the escaping, every shape in
this table came back truncated after an update, including the plain
one-paragraph record whose replacement holds a blank line.

The test asserts values only: the updated record as a second engine reads it
off disk, and the superseded content as Revisions returns it.

* style(memory): apply the linter's gofumpt form to the Update revision error

golangci-lint's bundled gofumpt wants the closing paren of that wrapped
fmt.Errorf on its own line; the standalone gofumpt binary the fmt task uses
(0.7.0, a different version) leaves it as it was. The text is unchanged from
before the engine fixes in this branch, so this is pre-existing drift inside a
function these commits already touch -- taken so the file is clean under both
formatters rather than only one.
2026-09-19 11:13:21 +10:00
sam 8b1aa44f6c feat(workintake): authorize review reactions by owned PR 2026-09-19 10:46:18 +10:00
sam 081fac96dc fix(config): stop a partial overlay clearing the fields it omits
An overlay or dashboard PATCH naming only some fields of a map-valued config entry
silently zeroed the rest. For [services.<name>] that meant losing the bearer token,
so changing a State Store endpoint dropped its authentication:

  start   Services{"state": {Target: "127.0.0.1:50051", TargetToken: "secret"}}
  overlay {"services": {"state": {"target": "10.0.0.5:50051"}}}
  result  Services{"state": {Target: "10.0.0.5:50051", TargetToken: ""}}

Cause: ApplyOverlayValues did yaml.Marshal(overrides) then yaml.Unmarshal into cfg.
yaml gives a struct FIELD the field-level treatment the doc comment promises, but a
map VALUE is decoded into a fresh zero value, so an entry the overlay only partly
addresses loses every field it does not name. That contract was written when
Services was a struct; making it map[string]ServiceConnection changed the behaviour
underneath it.

Fix at the producer: fold the fields the overlay omits forward out of the entry cfg
already holds, so the decoder is handed a complete entry. Recursion follows the
destination type, so the promise now holds for every map of structs -- services,
image.hosted, image.local and providers -- in one place instead of at each field's
reader. Struct field-level merge, scalar-valued map-key merge and slice replacement
are unchanged, and overrides is read, never mutated.

The regression test also converts to the map form, which is what unbreaks the build:
apply_test.go still built Services with the struct form after services.go became a
map, and because the two live in different files the breakage merged cleanly. There
is no Go build or test CI, so nothing compiled it.

Red: four cases failed with the token, the target and the sibling fields zeroed.
Green after the fold.
2026-09-19 03:07:14 +10:00
sam b332f03175 Merge pull request #878 from samcharles93/feat/attempt-attribution
Feat/attempt attribution
2026-09-19 02:48:25 +10:00
sam 080dd1d90b Merge pull request #879 from samcharles93/fix/docs-ci-and-overlay-coverage
fix(ci): unbreak the docs workflow, and cover nested config-overlay merge
2026-09-19 02:41:37 +10:00
sam 58a5ec9058 fix(ci): stop the docs workflow failing on a missing pip manifest
.github/workflows/docs.yml runs actions/setup-python@v6 with cache: pip. That
option hashes a requirements.txt or pyproject.toml to build the cache key and
fails outright when neither exists:

    ##[error]No file in ... matched to
            [**/requirements.txt or **/pyproject.toml]

This repository is Go and has no Python manifest, so the step failed as the
first step of every run triggered by the workflow's pull_request paths: docs/**
filter. mkdocs build --strict therefore never executed: docs changes were not
being verified, and every docs-touching PR was reported as a build failure. It
only surfaced on #878 because #878 touches docs/architecture/; PRs that do not
touch docs report no checks at all.

Drop cache: pip. The only Python use is a single `pip install mkdocs-material`,
so there is nothing worth caching.
2026-09-19 02:24:37 +10:00
sam 7b7401d98b fix(nats): bound the ARCHIE_REACTIONS stream by age
Review fix for PR #875. ARCHIE_REACTIONS is provisioned under LimitsPolicy,
which retains every message until a limit is reached. With no limit configured,
acknowledged reactions accumulated without bound and continuous traffic could
grow the stream until it exhausted disk.

nats.Config gains MaxAge *time.Duration, using the same pointer form as
Retention so "unset" (nil) is distinguishable from an explicit zero. It is
applied to the stream config only when set, so every existing caller keeps its
current behaviour.

ARCHIE_REACTIONS sets MaxAge = 24h. A reaction is a producer-only wake event
whose authoritative state is re-read at pass time, so a dropped reaction delays
work until the next poll (default 60s) rather than losing it; 24h tolerates a
consumer down for a full maintenance window and matches the package's existing
DefaultInactiveTTL.

ARCHIE_TASKS is deliberately left unbounded. It uses WorkQueuePolicy, which
discards a message on acknowledgement, so its size is bounded by outstanding
work; a limit there could discard a task that has not been delivered yet, which
is a worse failure than the disk risk. The table asserts that non-change.
2026-09-19 02:16:38 +10:00
sam 9e8427bc8e feat(nats): provision the ARCHIE_REACTIONS fan-out stream
archie.reaction.* had no matching stream. bootstrap.go:501-502 bound only
ARCHIE_TASKS to archie.task.> under WorkQueuePolicy, so any publish to a
reaction subject made js.PublishMsg fail and the publisher saw a publish error
-- the reason a previously landed reaction producer would have answered HTTP 500
on every webhook delivery.

ARCHIE_TASKS must stay WorkQueuePolicy: under it a second consumer on an
overlapping filter subject silently receives nothing, which is the trap recorded
in CLAUDE.md. Reactions need the opposite property, so they get a second stream:

- workintake.SubjectReactionWildcard ("archie.reaction.>")
- nats.DefaultReactionStreamName ("ARCHIE_REACTIONS")
- nats.FanOutRetention(), a pointer helper so composition can select
  LimitsPolicy without importing the NATS SDK, since nil Retention already means
  WorkQueuePolicy and the zero value (LimitsPolicy) is a valid explicit choice
- a second nats.Connect in connectNATS carrying that stream, closed on failure
  and registered for shutdown cleanup

Red-then-green: TestConnectNATSCreatesReactionStream failed with "publish to
archie.reaction.* = nats: no response from stream, want it to land in
ARCHIE_REACTIONS" before the stream existed.

Step 1 of docs/prds/pr-review-remediation.md (bead archie-core-8li9.1).
2026-09-19 01:51:07 +10:00
sam 6bdbd5e31e feat(config): key services by name through a registry
Services was a struct with one field per service, so each service's name was
stated three times: the field, the loader's defaulting branch, and every
consumer's field access. Adding one meant editing all three, and nothing
stopped them disagreeing. The drift was already there: stateStoreResolvedToken
and gatewayResolvedToken were identical apart from the env var each hard-coded.

Services becomes map[string]ServiceConnection, and RegisterService declares a
service once with the five fields it has. Everything that differs between
services is now an argument rather than a branch elsewhere: an empty registered
target means the operator must supply one, which is how services.state.target
stays required while the gateway's is defaulted, and the registered token env
var replaces the two copied resolvers with Services.ResolvedToken.

Defaulting iterates the registry, so a new service needs no loader edit.
Consumers read Services.Get(name).

Keeping a typo'd section reported took explicit work. [services.gatway] used to
land in toml.MetaData.Undecoded() because the struct had no such field; a map
decodes any section, so it would have loaded silently and left the daemon
dialing the default. unregisteredServiceKeys asks the registry which names are
real, and the test for it fails with an empty UnknownKeys when that check is
removed.

The TOML shape is unchanged, so no deployment profile, config.example.toml or
install.sh edit is needed.

Refs: archie-core-u23g
2026-09-18 23:32:36 +10:00
sam e0beb68978 feat(config): let the State Store and Gateway listen addresses be configured
Both gRPC listeners took their bind address from a -listen flag default alone,
with no config key. The port a service binds and the port its clients dial were
therefore two separate facts -- the flag default, and [services.<name>].target --
with nothing keeping them in step.

On a host that already owned the default the effect read as a code fault rather
than a port collision: the store died with "bind: address already in use" while
every client reported "state store: connection error ... error reading server
preface: EOF". Cockpit binds 9090 by default, so this is not hypothetical.
Verifying the dashboard on such a host needed a hand-written overlay plus a
matching -listen, two things to keep in step by hand.

- [services.state].listen and [services.gateway].listen now name the bind
  address. Precedence is -listen, then config, then a default applied by the
  loader. An empty result is a startup ERROR, never a fallback: net.Listen reads
  "" as "any free port", so defaulting there would bind the service somewhere no
  client looks and still report success.
- Both flag defaults become "" so configuration can win, mirroring how archie-ui
  already prefers [web].listen over its flag.
- Gateway listen defaults to the same constant as the gateway target, so a
  default deployment's client and server cannot drift onto different ports.
  [services.state].target still has no default: an empty target stays a startup
  error in the daemon and gateway, which no longer serve a local store.
- ServiceConnection gains the field. Its doc now says clients read only Target
  and TargetToken, and Listen belongs to the process that owns the service --
  the previous wording claimed the struct starts no listeners while carrying one.
- deployments/dev.toml states each port once, beside the target it must match, so
  `task dev` is retargetable on a host that owns 9090 by editing one overlay.

Tests. Unit: config defaults, explicit values, and precedence. Process: the
existing real-process smoke test now takes its listen from configuration instead
of -listen, so the end-to-end wiring is proven by the test that already starts
the binary, with no extra process. Both were verified load-bearing by mutation --
inverting the precedence fails the unit test, and making the app layer ignore
config fails the process test with "state: no listen address".

Verified: task check exit 0. Also documented in config.example.toml, the
configuration architecture table, and the `task dev` summary.
2026-09-18 19:15:02 +10:00
sam 2ef962912a feat(tasks): attribute events, changes and config to the attempt that produced them
The events table had no attempt column, so two attempts of the same task were
indistinguishable: a retried task's stage events interleaved into one timeline and
nothing could say which run produced a given result.

- events gains an `attempt` column through the existing presence-gated
  migrateTasks loop rather than a version bump, so user_version stays 3 and an
  existing archie.db is migrated in place. A legacy-database test builds the
  pre-change schema and proves both that existing rows survive unattributed
  (attempt = 0) and that a new value round-trips.
- Event.attempt = 11 and ReadTaskLogRequest.stage = 9 are field additions to
  existing messages: no new RPC, no new table, and no change to any canonical error
  message that crosses the state-store wire.
- Every task event the workflow engine emits now carries the attempt. Producers
  that have no attempt yet (task_queued, binding_dispatch_failure) stay
  unattributed rather than guessing one.
- Changed files are captured as a durable changes_captured event at commit and push
  time, reusing the existing three-dot diff primitive: paths from FilePatches(),
  counts from Stats(), base_sha recorded as the merge base actually diffed rather
  than the base tip, a 200-file cap with totals kept complete. Consumption goes
  through an unexported optional interface so workflow.Trees -- the yaegi capability
  surface for interpreted stage plugins -- is not widened. A capture is also taken
  once the pull request exists, so the capture can link it.
- The per-attempt configuration is captured as a durable config_captured event
  carrying the non-secret config.TaskConfig, instead of a new table plus two new
  state-store RPCs: events are already this system's per-attempt provenance stream,
  so a table would have given one feature two provenance mechanisms. Redaction is
  proven against canary secrets. Measured payload: 539 bytes typical, 2333 bytes at
  20 models.
- Captures carry a schema marker and are refused unless their key set matches the
  declared payload shape, so drift surfaces as "recorded but could not be read"
  rather than as a capture of zeroes.
- Three read endpoints serve the attempt rail, the changed-file captures and a raw
  debug view, mapping to UI-owned response types rather than dumping store types.
- The task-log read gains a stage filter, bound to the stage logger for that
  stage's own lines; agent and tool output records no stage.

A task-scoped grant may already insert events on its own task, so this provenance
is exactly as trustworthy as the rest of that task's event stream. That limit is
documented rather than papered over.
2026-09-18 18:32:59 +10:00
sam a484693f11 refactor(logging): pass the caller's context to TaskRegistry.Write
Write built its own context.Background() for the handler call, which meant
any new caller with a context to offer could not hand it over. Taking the
caller's context is what the next commit's park-reason writer needs, and it
is what contextcheck requires of a daemon-side writer.

No behaviour change: the context is used for the handler's own attribute
extraction and never consulted for cancellation -- a park is recorded at
exactly the moment a run's context has been cancelled, so a cancelled
context must not suppress the entry.
2026-09-18 01:49:42 +10:00
sam c6dcfd6671 fix(staterpc): never decode a collection field to a nil slice
Every client method that decodes a repeated proto field goes through
mapValues, which returned nil for nil input. Protobuf decodes an empty
repeated field to nil, so an empty collection crossed the wire as JSON null
rather than an empty list: the field's type depended on whether it happened
to have contents, and the wire path disagreed with the local path, which
normalises its own slices.

This is the nil-slice class the contract-conformance audit deferred, arriving
by the route its own revisit condition names -- a handler that can produce
nil. The audit checked the webui handlers (clean, as it said) and did not
cover the wire mapping added afterwards, so the regressed field was the new
one: ReadTaskLog's components, observed as "components": null from the live
dashboard.

mapValues now returns []B{} for empty input, which fixes the class once at
the mapping helper rather than at each of its 36 call sites -- none of them
compares the result to nil, so nothing depended on the old behaviour.
TaskLog also normalises components, which is read straight from the proto
rather than through mapValues.

Red first: the test drives the existing local/grpc matrix and asserts the
marshalled JSON, so it fails on a null collection without depending on which
path is under test. Both collections are exercised empty -- entries present
with components empty, and a log file that exists with no lines -- because
the first shape alone leaves the entries side untested. Verified by mutation:
restoring the nil return fails it on "entries marshalled to null".
2026-09-17 14:58:14 +00:00
sam e338473414 Merge pull request #846 from samcharles93/feat/835-disk-readiness-probe-watches-only-work-dir-so-a-full-root-fi
Disk readiness probe watches only work_dir, so a full root filesystem goes unreported (archie)
2026-09-17 22:39:15 +10:00
sam 4ef56edd77 Merge pull request #848 from samcharles93/feat/state-store-task-log-download
feat(staterpc): read task logs through the State Store and make them downloadable
2026-09-17 22:38:16 +10:00
sam 21d8250cde feat(webui): read a task's log through the State Store and let it be downloaded
The dashboard's attempt-log panel answered every read with "task logging is
optional and was not enabled for this run". That message was false: logging is
wired unconditionally in the daemon, and the log files exist. The dashboard
process simply had no reader -- composition left TaskLogs absent because no
contract carried it -- so the handler saw a nil registry and reported the
disabled shape, which the page rendered as a claim about configuration.

internal/logging now answers whether an attempt has a log at all: Tail sets
Result.Found, so "this attempt has no log" is a value rather than an empty
list indistinguishable from a read that never happened. Its registry gains
TaskLog (a decoded page) and TaskLogContent (the file verbatim, streamed,
because a log is unbounded input), and an absent reader reports
ErrTaskLogsUnavailable rather than an empty page -- a process that cannot read
logs must not implicate the attempt.

Those two answers travel the State Store contract as storecontract.TaskLogStore
(ReadTaskLog, StreamTaskLogContent) and are rehydrated on the client by
(Unavailable, canonical message), so archie-ui composes the reader from its own
store client instead of opening another process's files. The daemon's registry
is wired as the server-side reader in the State Store process only.

On top of that read, GET /api/tasks/{id}/logs/download serves the attempt's log
as a Content-Disposition attachment named task-<id>-attempt-<n>.log, and the
panel offers it. The JSON route keeps its shape and gains found, so the page
states the attempt has no log instead of blaming a setting.

Task log files stay owned by their directory's owner: archiving still removes a
task's log directory, and that removal is the daemon's, not the dashboard's.
2026-09-17 12:12:43 +00:00
sam 2cd796a3a3 feat(gateway): persist the channel-native sender id on messages
Slice 1 of docs/prds/memory-engine-unification.md: persist the
participant, so a later curator can address agent-user memory scope.

- messages gains sender_id: schema, ALTER migration (idempotent via
  duplicate-column detection), both inserts, and both history selects.
  Pre-existing rows default to empty.
- curator.SessionSummary gains AgentID and ConversationMessage gains
  SenderID; the sessioncurator adapter fills them from the session's bot
  identity and the stored record.

The memory contract is untouched and nothing consumes the new fields
yet. Deriving a session's participant from its user-role messages, and
failing closed on ambiguity, stays with the curator rewrite.
2026-09-17 22:06:54 +10:00
sam 51a05c0f33 fix(readiness): keep required status when an optional target shares a path
Targets are deduplicated by path so a shared filesystem is probed once, but
the first target seen won. With root, then an optional /var, then a required
data path that resolves to /var, the optional target is the one that survived
-- so its rules were applied to a path the daemon requires. An absent /var is
skipped as "optional absent", and with another healthy target present the
probe reported OK while a required path was unavailable.

Deduplication now keeps the strictest requirement any target places on a
path, and its name, so the failure is both visible and correctly attributed
rather than reported against the optional target that happened to come first.

Red first: the new test drives the reported scenario through an injected
statfs and failed with `status = "ok", want degraded` before the fix.
Verified by mutation -- restoring first-wins dedup fails it again.
2026-09-17 11:48:23 +00:00
samcharles93 8378c92739 Disk readiness probe watches only work_dir, so a full root filesystem goes unreported (archie)
Implements #835
2026-09-17 11:34:38 +00:00
sam f25dd1e867 refactor(config): delete the two decoded-but-unread policy fields
MemoryConfig.SessionTTL and ToolPolicy.ParallelExecution were decoded by
config loading and read nowhere, so setting either silently did nothing.
Removing them routes both keys into the unknown-key detector, which names
them in a boot warning instead. A live-looking knob wired to nothing is
the worse state, and the sibling memory.provider already earned a loud
load-time rejection for exactly this.

Also drops the config.example.toml line that shipped session_ttl into
every first-run config.toml via configtemplate.go, and the commented-out
dir_test.go assertions that would have re-created both fields under that
file's "make these pass without modifying them" header.
2026-09-17 00:32:02 +10:00
sam 0020867b3e docs(skill): drop comments that named the deleted skill.Discover
Lane F deleted skill.Discover/skill.DiscoverPlugins/skill.Skill (no
production caller) in 6f4e420, which left three comments in packages it did
not own describing or contrasting behaviour against a symbol that no longer
exists -- a dangling reference is the same drift class as an unreachable
code path, just cheaper to fix.

The behavioural statements are kept; only the dead names go. internal/skill's
Catalog skips a SKILL.md it cannot read, while this curator's store reports
the per-skill failure as an Action, so the test's rationale still holds
without naming the removed function.

Comment-only: no behaviour change, task check exit 0.
2026-09-15 13:17:54 +10:00
sam fb945fec41 fix(taskactions): resolve max_retries from the task's owning identity
WIRE, not delete: per-repo max_retries is a configured field with a real
consumer (the retry gate), and it works for the legacy single-identity
[[repos]] list. In a multi-identity deployment repositories live under
[[identities.repos]] and there is no global [[repos]] at all, so the resolver
never saw the override and a parked task was killed at the global cap rather
than the configured one.

MaxRetries now consults the identity named on the task before the legacy list,
then falls back to the global cap. Identity name is the only thing that says
which repository list owns a persisted row.

Red-then-green: TestMaxRetriesResolvesFromTheOwningIdentity gained the
multi-identity cases; "identity repo override wins over the global cap" failed
with `MaxRetries = 3, want 7`. Two neighbouring cases keep the fix honest:
another identity's override must not leak, and the legacy path must still
resolve. Mutation-verified by deleting only the identity lookup, keeping the
extracted helper and the API intact.
2026-09-15 12:59:06 +10:00
sam 8fdddb0463 merge: lane E (tools-media-1, transport-nats-2, release-1) 2026-09-15 12:55:07 +10:00
sam 681824ccd0 merge: lane D (curator-1, staterpc-2) 2026-09-15 12:53:32 +10:00
sam ddeb654f43 fix(scheduling): record a completed run so a recurring job stops re-firing every tick
WIRE, not delete: cronstore.Store.MarkRun is the documented run-completion
step ("once the ticker has dispatched a job, the runner ... calls MarkRun to
keep next_run fresh") and the scheduling engine is now started at boot, but
nothing in production called it. Due returns every job whose NextRun is at or
before now, so with NextRun frozen in the past a job with a 1h interval was
handed to the runner on every tick.

The delivery Router is the single Runner the engine is given and, per this
package's own doc, the only place that knows a job store exists, so it is the
producer-side site. RouterStore (SpecLookup + RunRecorder) widens only what
the Router takes; the per-kind runners keep the narrow SpecLookup, and
NewRouter's signature is unchanged, so no call site moved.

Only a successful run records, matching the store's own contract that an
interval or cron schedule fires "after the last successful run". A one-shot
schedule reports ErrScheduleUnsupported -- it has no recurring next run by
definition -- which is swallowed rather than returned, because propagating it
would make the engine emit KindJobError for a job that actually succeeded.

Red-then-green: TestRouterAdvancesTheScheduleAfterASuccessfulRun asserts
LastRun is set and NextRun moved past now, and failed on both counts before
the change. TestRouterToleratesAOnceScheduleWithoutANextRun pins the
sentinel half. Mutation-verified by replacing only the record call with
`return nil`, keeping the API intact.
2026-09-15 12:53:02 +10:00
sam a85ba42870 fix(servicediscovery/nats): validate config after defaults
Root-cause fix: Connect validated the raw config before withDefaults, so
the Heartbeat<TTL guard was skipped whenever Heartbeat was unset and a
custom TTL below DefaultHeartbeat passed as a silently expiring registry.
Apply defaults first so the documented invariant is enforced. No
WIRE/DELETE decision applies: this is a producer-ordering bug, not
unreachable code.
2026-09-15 12:51:17 +10:00
sam 42adee2f87 fix(staterpc): preserve marshal_error marker in eventDataJSON
WIRE: docs/prds/state-store-contract.md §11 requires the remote wire and the
local adapter to behave identically. The local adapter (internal/store/events.go)
records a {"marshal_error":%q} marker when json.Marshal of event Data fails,
while eventDataJSON silently returned "" on the same failure, emptying the
event's Data on the wire. Carry the same marker so an unmarshalable Data
value survives the hop with the local adapter's fidelity.
2026-09-15 12:50:57 +10:00
sam e84aded3fb refactor(state): delete the dead IncrementRetryCount RPC (breaking change, accepted)
IncrementRetryCount had zero production callers -- only tests and the
staterpc plumbing referenced it. taskactions.Service.applyRetry already
uses the atomic RetryTask exclusively, which increments retry_count and
requeues in one transaction.

Unlike ListCaptures/ListUndispatchedCaptures (kept as declared-unconsumed
compatibility surfaces because a rolling deploy could have an old
client/server pairing mid-flight), this repo has one production consumer
(see AGENTS.md's solo-project note), so the breaking removal is accepted
outright rather than deprecated-and-kept. Deleted the Go method, the
storecontract interface entry, the staterpc server/client methods, and the
RPC/messages from state.proto, regenerating the .pb.go/_grpc.pb.go pair.

Test fixtures that used IncrementRetryCount as a setup helper now drive
retry_count through the real path instead: queries_test.go bumps it via
direct SQL (same package, no contract to route through), and
api_tasks_actions_test.go now loops RetryTask+Transition, which exercises
production code instead of a deleted one.
2026-09-15 11:24:57 +10:00
sam 24e94828a3 refactor(staterpc): delete the superseded single-token auth interceptor
UnaryTokenInterceptor/TokenValidator/checkToken (auth.go) and
constantTimeTokenValidator (state_store.go) had no production caller:
server auth on the bridge-address listener is wired through
TaskGrants.UnaryInterceptor/StreamInterceptor, which replaced this
all-or-nothing token check with per-task scoped grants. Removed the
dead cluster and its now-pointless tests; TaskGrants' own auth
rejection paths are already covered in task_grants_test.go.
2026-09-15 10:18:25 +10:00
sam 8139c0a3bf feat(configuration): report config keys no decode target consumed
decodeTOML discarded toml.DecodeFile's metadata, so a misspelled or
renamed key parsed, validated and silently did nothing. Document.UnknownKeys
now carries the dotted key paths a loaded file's decode targets left
undecoded.

One file feeds two decode targets (&doc.Config and &doc.Scheduling), so a
key legitimately owned by one shows up as undecoded in the other -- a key
is only reported when both leave it undecoded (unknownKeys intersects the
two sets), which is what keeps a valid [scheduling] block from being
misreported as unknown on every config that uses it.

YAML config sources are unchanged for now (no equivalent metadata hook in
the yaml decoder); this covers the TOML path, which is what
config.example.toml, every shipped deployments/*.toml profile, and
install.sh's generated config.toml all use.
2026-09-15 05:07:54 +10:00
sam a856e5ac80 fix(archied): disable a provider whose credential is missing instead of refusing to boot
configuration.md's startup policy -- "a missing credential disables a capability.
Only an invalid config stops the daemon" -- names this exact case: "Apply the same
reasoning to any future capability credential, LLM keys especially: start, report
the capability as unavailable, let the operator see it." The code did the
opposite. resolveProviderMap returned an error when a provider's api_key could not
be resolved, and again when it resolved empty, so configuredSecretRegistry failed
and the daemon would not start. A provider nobody referenced still blocked boot,
because the walk covers every entry in cfg.Providers rather than the ones [models]
uses.

The forge path was fixed to degrade for precisely this reason (resolveForge,
guarded by TestResolveForgeDegradesInsteadOfFailing); this is its provider
counterpart. A credential that cannot be resolved now disables that provider with
a warning naming it and why, and the daemon starts.

Disabling clears the reference rather than leaving it. webui's ProviderView
derives Configured from api_key or APIKeyEnv, so an unresolvable reference left in
place would advertise a provider nothing can use -- the "looks configured and is
not" shape this codebase keeps having to remove. Consumers already treat an empty
APIKeyEnv as no credential (agentexec/runtime skips such a provider, the daemon
exports nothing to the agent container), so clearing it is the whole disable.

The other half of the policy is kept: a reference naming only one of engine and
key is a config error, not a missing credential, and stays fatal.
applyForgeDefaults only ever builds a ref with both halves, so a half-named one can
only come from a hand-edited config.

This also defuses the trap 027dc99 introduced. archied setup can now write a
provider reference to any engine (-provider-secret-ref bws:OPENAI_API_KEY), and on
a machine without that engine the config used to load clean and then brick the
daemon. It now starts with that provider disabled and a warning.

The comment on templateDefaultActiveProvider described the old fatal behaviour and
is corrected rather than left to mislead the next reader.

All three behaviours are mutation-verified: returning an error again, accepting a
half-named reference, and disabling without clearing the reference each fail the
corresponding test. The red was an assertion failure, not a compile error -- the
logger parameter was threaded through first so the new test could be run against
unchanged behaviour.
2026-09-15 03:28:35 +10:00
sam 027dc9925b feat(configuration): let setup reference a secret without carrying its value
A parameterised install could not configure a cloud provider without a terminal:
the only way a secret value entered was the prompt, so an unattended run was
limited to the keyless self-hosted provider. The missing piece is the reference
form.

A reference is not a secret. An engine name and a key name are not sensitive --
which is exactly why a reference may be a parameter while a value may not -- so
setup gains -forge-secret-ref, -provider-secret-ref and -telegram-secret-ref,
each taking engine:key, parsed by one new setup.ParseSecretRef. When a reference
is given the step writes it to TOML and asks nothing, and setup neither reads nor
stores the value, because the value is not its to hold. That is the difference
from the prompt path, where the operator supplies a value and setup writes it
into the env engine and references it there.

Engine names are identifiers, not paths: the shipped engines are registered as
age, sops and vault, so a reference like vault/openbao:KEY is refused rather than
written into a config that could never resolve it.

Two refusals guard the silent no-op. A reference naming a key for the keyless
self-hosted provider, or for a forge that is disabled, would sit in the config
looking configured while nothing read it -- the failure shape this flow exists to
remove -- so both are refused rather than written.

telegram keeps writing token_env for the prompt path, because that form names an
env var and the prompt path is the env engine; a supplied reference may name any
engine, which token_env cannot express, so the reference path writes the
preferred token = {engine, key} form. Noted as the one asymmetry.

Also folded stepCloudProvider's key handling into cloudKeyEdit and cloudModel.
That was a lint fix for nested blocks, but the split is the honest shape: the two
branches obtain a key edit by different means and then resolve a model the same
way.

Every new test was mutation-verified by reintroducing the absence: ignoring the
forge reference, making ParseSecretRef validate nothing, and dropping the keyless
-provider guard each fail the corresponding test.
2026-09-15 03:23:08 +10:00
sam 24ba2ce609 fix(configuration): stop letting a secret value be set from the command line
archied setup took -forge-token, -provider-api-key and -telegram-token, each
carrying a secret VALUE. That is wrong twice over: the value lands in shell
history and in every process listing, and it is a second way to set a secret
that no secret engine knows about.

The codebase already holds the opposite invariant, and had before this landed:

- configuration.md:423-425 states it directly -- "setup writes the reference to
  TOML and the value to the secret engine".
- No CLI flag anywhere in the repository accepted a secret value; these three
  were the only ones.
- install.sh already does it correctly: read_secret prompts without echoing so
  the value never reaches scrollback, and set_env_key writes it into the systemd
  EnvironmentFile, which is the env engine's backing store.

So the SecretSink half was right -- writing the value to the env file IS writing
it to the secret engine, for the env engine -- and the prompt path was right.
Only the three value-carrying flags were wrong.

Removed: Params.ForgeToken, Params.ProviderAPIKey, Params.TelegramToken and the
flags that fed them, along with the parameter plumbing in step_forge,
step_provider and step_chat. Those steps now always read the secret from the
prompt surface. Params gains a comment recording why it deliberately has no
secret-valued field, so the absence is a decision rather than an oversight.

TelegramUserIDs stays: an allowlist of user IDs is access policy, not a
credential.

Two tests pinned the wrong behaviour and are replaced rather than deleted:

- TestRunParams_ProviderAPIKeyNeverPrompts asserted a secret could arrive as a
  parameter. It becomes TestRunParams_SecretsComeOnlyFromThePrompter, which
  supplies every non-secret answer through Params and scripts only the secret
  reads, so it fails if a secret ever comes from anywhere else.
- TestRunSetup_DefaultsWithForgeTokenWritesEnv exercised the removed flag. It
  becomes TestRunSetupRejectsSecretValueFlags, which asserts all three flags are
  rejected with exit 2 and that neither a config nor an env file is written --
  verified by re-adding -forge-token and watching it fail.

The unused rejectingPrompter went with them; its remaining replacement uses a
scripted prompter instead.
2026-09-15 03:15:53 +10:00
sam ee608d81fb feat(configuration): add archied setup, the caller the first-run flow was built for
docs/architecture/configuration.md decides that "install.sh must stop generating
config.toml; archied setup owns it", and gives the reason: the installer
hand-rolling TOML already drifted from the schema once, and "the durable fix is
that the code writing the config is the code that reads it". setup.Run,
tomlwrite, configtemplate.Example and terminalprompt were all built for that and
had no production caller between them. This adds the caller.

setup.Run's own doc comment states the obligations, and they are the shape of
installValidatedConfig: render the edits, prove the result loads through a real
configuration.Loader, and only then install the file and commit the secrets. A
validation failure returns before either write, so a bad config leaves neither a
config nor a secret on disk. The rendered bytes are proven from a temp file, so
a load failure cannot have already installed anything. The config is written
0600 via same-directory temp and rename.

Parameterisation is typed, not matched against prompt text. A flag-to-prompt
mapping would have to match strings, which is the brittleness removed in
9a4f11c, so setup.Params is consulted before each question instead and the
prompts are only reached for answers nothing supplied. Run stays a wrapper
passing the zero value, so the package's existing tests keep passing untouched.
No environment variable is consulted anywhere: flags and prompts are the only
two ways an answer arrives.

--defaults is a known-valid unattended baseline, not "each prompt's own default".
That cannot work, and the test proves it: the provider step's first option is
OpenAI, whose model prompt hard-errors on an empty answer. The baseline therefore
covers every required answer and leaves optional ones to answer themselves as
"skip" -- ollama is the only keyless provider, so an unattended install boots
without needing a secret.

Two gaps closed after review rather than left as accepted work:

- A cloud provider was not configurable without a terminal, because the provider
  API key had no typed parameter: unattended setup could only ever pick the
  keyless self-hosted provider. Params.ProviderAPIKey and -provider-api-key fix
  that. The key goes to the env file and TOML carries only the secret reference.
- The write order was unpinned. Inverting installValidatedConfig's two writes --
  committing secrets before the config -- left every test green, so nothing held
  the invariant the SecretSink contract states. The new
  TestInstallValidatedConfig_ConfigWriteFailureLeavesNoSecret fails on that
  inversion; the existing validation-failure test does not cover it.

Also fixed three lint failures in the first pass (gocyclo on RunSetup, and two
ineffectual `choice` initialisers): RunSetup's flag surface, the flag-to-Params
mapping, the prompter choice and the existing-config read are now separate
functions, which is the real fix for the complexity rather than a suppression.

The setup flow no longer hand-rolls its own answer format: -telegram-user-ids
reuses setup.ParseTelegramUserIDs, the same parser the prompt path uses, so the
allowlist cannot be parsed two ways.

Found by an adversarial package sweep (temp tracker: config-1,
repo-root-deployment-2, .local/issue-tracker/). Plan of record:
.local/issue-tracker/design-archied-setup.md. install.sh still generates its own
config; that is phase 2 and is not claimed here.
2026-09-15 02:52:24 +10:00
sam dd3ae02472 fix(scheduling): respect configuration ownership 2026-09-14 23:25:09 +10:00
sam 67ce0d1c4d feat(cronevents): adapt the events bus to scheduling.Sink
A scheduled run has no task, no chat turn and no workflow behind it, so its
emitted events are the only trace it leaves -- and the engine's Sink is the
emitter shape of a bus it deliberately does not import. Something has to turn
one into the other; this is that seam.

  - Sink(mode, bus) returns a sink over the bus for "bus", and nil for
    "none"/absent or a nil bus. Nil rather than a no-op sink keeps
    "emission is off" visible to whoever reads the composition, and it is
    the configuration the engine documents as valid: it still runs, it is
    just unobservable.
  - The published event carries the engine's own Kind/Detail/Data and no
    task-scoped identity. A cron run has none, and borrowing the emitting
    daemon's would attach it to an unrelated task's timeline.
  - An unrecognised mode is an error; the loader rejects it first, but a
    hand-built config must not silently disable observability.

TDD: the sink tests were written first and failed on assertions (Sink("bus")
returned nil; an unknown mode was accepted), not on compilation. The
end-to-end test drives a real engine on an injected clock and subscribes to
the bus, so the Emit -> Publish path is proven rather than assumed -- no
sleeps, no wall-clock deadlines.
2026-09-14 10:51:47 +00:00
sam e9cc3076e4 feat(configuration): translate [scheduling] into scheduling.EngineConfig
The engine needs a fully-resolved configuration, not zero values it has to
second-guess. Add the translator the wiring slice then hands to
internal/app/archied:

  - TranslateScheduling resolves unset tunables from the domain's own
    Default* constants rather than a second copy here, so the two cannot
    drift.
  - A negative tunable is reported as a startup error wrapping
    ErrInvalidInput, matching validateCapture and validatePollInterval.
    Clamping was rejected: it runs something other than what the operator
    wrote, which is how a misconfiguration stays invisible.
  - Events and Clock stay nil. This package owns no bus and no clock --
    attaching them is composition -- and the engine documents a nil sink as
    a valid configuration: it still runs, it is just unobservable.

applySchedulingDefaults and validateScheduling keep defaulting and
validation separate, the split applyDefaults' own doc explains: the load
path applies both, Validate applies only the latter.

TDD: the translator tests were written first and failed on assertions
(Interval = 0s, want 1m0s; negative values accepted silently), not on
compilation.
2026-09-14 10:51:41 +00:00
sam 82b28cd7f4 feat(crondelivery): add chat, workflow and routing cron runners
The ticker engine had no work to dispatch. Add the delivery half as a new
infrastructure package with three scheduling.Runner implementations:

  - ChatCourier sends a job's payload text to its target chat.
  - WorkflowTask submits a job as a unit of work.
  - Router dispatches on the job's kind and is the single runner the engine
    is given, so the engine stays unaware that more than one delivery exists.

Nothing here talks to a channel, a broker or a forge. The outbound send is
an injected Courier function and work intake is a narrow TaskSubmitter
interface, so which transport carries a message stays the deployment's
decision rather than the cron's, and the package is testable without a
network.

An unrecognised kind is refused at the gate, not failed: it emits
KindJobError with phase "dispatch" in the engine's own data shape and
returns nil, so a misconfigured job is visible without being reported as a
job that tried and broke. A nil sink still refuses, just unobservably.

Every runner returns the context error on a cancelled run and produces no
side effect, so the engine's bounded shutdown drains rather than waiting
out the per-run timeout.

Tests exercise a real cronstore on a temp path, including the epic's
headline daily-status scenario end to end from a persisted due job through
the router to the delivered message.
2026-09-14 09:54:41 +00:00
sam 7b13751783 feat(cronstore): add job kind discriminator and bump schema to 2
Jobs need a discriminator for the delivery router to dispatch on, and
JobSpec had none. Add a top-level Kind field with values "chat" (default
when empty) and "workflow".

The empty value is load-bearing: every job written before this field
existed carries no kind and must keep delivering as a chat message, so an
empty Kind resolves to "chat" rather than being rejected.

DisallowUnknownFields means this additive change must bump schemaVersion,
so it goes 1 -> 2. A version-1 file still loads — Kind reads back empty —
which a test pins, because the alternative is that every existing cron job
stops firing on upgrade.

The store does not interpret the value: an unrecognised kind is persisted
verbatim and refused by the router, which owns the kind-to-runner mapping.
2026-09-14 09:54:41 +00:00
sam 408b8c6bff test(archied): stop racing state-store startup against a wall clock
TestRunStateStoreAcceptsEnvironmentToken wrapped the whole of RunStateStore --
SQLite schema init and the serve loop -- in a 200ms context deadline, so
whether it passed depended on how fast the disk was. Measured before the fix:
6 failures in 12 consecutive runs. A gate that fails half the time teaches
people to ignore it, and the failure it reported ('store: init schema:
context deadline exceeded') named a time budget, not a defect.

Cancellation is the assertion, so the test now cancels and then asserts:
the call must not return before cancellation (that would mean startup refused
the config, the thing this test exists to catch) and must return within a
generous grace once cancelled. The graces bound 'did startup reject us', not
'machine speed'. Measured after: 12 passes in 12 runs.

Also applies the canonical gofumpt and 'go fix' output to
internal/infrastructure/staterpc/dial.go, cmd/archie-ui/main_test.go and
internal/releaseupdate/check_script_test.go. task check runs gofumpt -w over
the tree, so leaving these unformatted made the gate rewrite unrelated files
mid-run and report dirty state for changes it had made itself.
2026-09-14 08:09:14 +00:00
sam 19062a702d refactor(webui): take configuration off the dashboard server
webui.Server declared Cfg *config.Holder with eight read sites, none of
which ran in the UI process: composition leaves it nil there. The field
survived because the daemon kept a webui.Server it never served, using it
to render the published ConfigView snapshot and to hold the readiness
probes (archie-core-ml30).

Split at the producer. LocalConfigView becomes webui.BuildConfigView, a
pure function over a ConfigViewInput the configuration owner assembles;
the DTO and its schema stay where the UI renders them. The daemon builds
that input from its own state (boot.configViewInput) and publishes the
result. Readiness reads the daemon's Holder, channel manager and chat
contract, and the registry it produces is a daemon field, not a field on
a server with no listener. Boot now owns one Holder instead of seeding a
second and replacing it in buildDaemon; operator retry caps read it too,
so they follow a reload rather than a boot-time snapshot.

Server loses Cfg, ConfigProvenance, LastReload, ConfigOverrides and
SetProvenance. The nil-guarded branches they fed went with them, each
checked first against what the UI process actually serves:

- chat.show_tool_calls and the /api/setup checklist were degraded, not
  dead. ConfigView gains a chat section carrying show_tool_calls, the
  operator name and whether a channel is configured, and both surfaces
  read the projection. The end-to-end suite asserts the checklist renders
  from the published snapshot.
- /api/logs takes its path as a process-local option. The history is on
  the daemon's disk, so a dashboard elsewhere cannot read it by learning
  the path; that route still needs a contract.
- handleChannels drops its config-derived fallback, superseded by the
  status manager, which already carries the descriptions and detail.
- trustForwardedHeaders keeps only the process's own setting, which the
  UI process already sets.
- resolveForge keeps the published-projection path it already had.

ConfigView.Editable moves to handleConfig, decided by whether this
process holds UpdateConfig. It describes the writer, not the document, so
the write contract (archie-core-j28m) makes the page editable by wiring
one field. The write handlers and their seams stay for it.

overlay.Store gains Keys: Snapshot nests dotted keys for decoding, so the
overridden-key list was publishing section names ("budgets") where the
dashboard marks and resets rows by stored key ("budgets.max_steps").

Multi-identity forge links move to archie-core-pv6t: forgeConfigForTask
resolved a task's forge from config.Config, had no reachable caller once
the holder was gone, and belongs in the projection rather than the
rendering package.
2026-09-11 23:57:38 +10:00
sam af2c8e1f41 fix(staterpc): attach the bearer token to State Store streams
The capture batch reads moved onto server-streaming RPCs (StreamCaptures /
StreamUndispatchedCaptures) to get past gRPC's 4 MiB unary cap, but the
client interceptor suite attached the bearer token to unary calls only:
Dial installed WithUnaryInterceptor and nothing for streams, while
TaskGrants.StreamInterceptor requires the administrative token. On every
token-protected (non-loopback) listener -- every container-mode deployment
-- the daemon's binding dispatch and the dashboard's capture inspector got
PermissionDenied from the stream interceptor, and the unary fallback those
reads replaced is no longer taken, so there was no working path.

- Add StreamClientTokenInterceptor, mirroring gatewayrpc's, and install it
  alongside the unary one in Dial, the single client dial path the daemon,
  the dashboard and the agent all use.
- Route agenttransport's State Store dial through staterpc.Dial, deleting
  the duplicate dial/fail-closed logic that installed only the unary
  interceptor -- a second place for this class of bug to hide.
- Regression tests: the grants harness now dials through the production
  Dial so client-side wiring cannot drift again;
  TestCaptureStreamsCarryTheAdminToken reads both list surfaces (5 MiB
  batch) through a token-protected listener;
  TestStateStoreRealProcessTokenProtectedCaptureStream does the same
  against the real binary; TestUpdateRejectsNilTask covers the Update
  guard left untested by the earlier finding fix.
2026-09-11 22:43:34 +10:00
sam da28ddb5e6 refactor(ui): sever the archie-ui binary from every daemon runtime package
Completes the objective deletion gate for the UI Service
(archie-core-8cda.5.6): go list -deps ./cmd/archie-ui now links zero
banned packages -- no modernc.org/sqlite, no internal/store, gateway,
channels, memory, secret, agentexec, tools, skill, yaegiutil, curator,
or the workflow engine. The closure shrinks from 68 to 26 archie-core
packages. Each cut is a producer-side move; no consumer-side defensive
shims.

- Store contracts (TaskStore family, value types, sentinels) moved from
  internal/store to internal/domain/storecontract -- the approved contract
  location per state-store-contract rev. 2d, and NOT internal/contracts:
  that tree is buf generate output and tools/proto.sh check deletes
  hand-written files there. internal/store keeps aliases; sentinel
  message strings are byte-identical and remain the wire contract.
- Chat contracts (ChatContract, sessions, turns, streaming, commands)
  moved from internal/gateway to internal/domain/messaging; gateway keeps
  aliases. ToolCallEvent rendering helpers become gateway functions
  (methods cannot live on alias types).
- Task vocabulary (Task, statuses, mid-run Store, Definition) split into
  internal/domain/workflow/task; the pipeline engine stays in package
  workflow with aliases, so daemon-side callers are unchanged.
- overlay.DeniedKeys moved to internal/infrastructure/configuration:
  webui referenced the constant and that dragged the overlay SQLite
  store in.
- webui.Server's concrete runtime handles become webui-owned view
  interfaces with daemon-side bootstrap adapters: channel lifecycle
  (internal/channels/status), MemoryStatus, CuratorStatus, and a
  SkillCatalog seam replacing the direct skill.DefaultRoots/CatalogRoots
  calls. Each degrades to an empty page when unwired.
- config.SecretRef moved into internal/config (config vocabulary);
  internal/secret keeps an alias and Registry.Resolve absorbs the
  zero-value short-circuit.

Docs: state-store-contract rev. 2d relocation record; ui-service-boundary
deletion gate marked satisfied; ARCHITECTURE.md package map rows.
2026-09-10 03:03:33 +10:00
sam 3699f51f9b refactor(store): move the producer-owned contracts to internal/contracts/store/v1
The deletion gate for the UI Service (archie-core-8cda.5.6) is objective:
go list -deps from cmd/archie-ui must show no SQL/store implementation.
internal/store's hand-written contract surfaces (TaskStore and its
component interfaces, CaptureStore, MappingStore, BindingStore,
BindingDispatcher, BindingTaskCreator, ConfigSnapshotStore) move to
internal/contracts/store/v1 (package storev1), alongside the value types
they carry (CapturedEvent, ConfigSnapshot, WorkflowStat, StageStat,
DayTokens) and the sentinel errors -- whose canonical message strings are
the wire contract and are byte-identical after the move.

internal/store keeps type aliases, so daemon-side callers (archied, the
daemon, staterpc's server side, taskactions) are unaffected. The
UI-facing packages switch to storev1 directly: webui, archieui,
captureintake, gatewayrpc, and staterpc's client/values -- which drops
internal/store and modernc.org/sqlite from the archie-ui binary's import
closure.

overlay.DeniedKeys moves to internal/infrastructure/configuration for the
same reason: webui referenced the constant and that dragged the overlay
package's SQLite store in.

The chat-contract and remaining runtime-handle decoupling continues under
the same bead.
2026-09-10 03:03:33 +10:00
sam 27bc40332f refactor(archied): cut the dashboard over to the archie-ui process
PRD gate 7 (docs/prds/ui-service-boundary.md): the daemon's in-process
webui.Run listener is deleted in the same change that wires the UI process,
so no fallback or second HTTP authority remains. The daemon's webui.Server
stays only as the configuration snapshot's renderer (archie-core-ymut): it
keeps the config Holder, provenance, reload status and the override list,
and drops the store surfaces, event broadcast, chat/task wiring that only
fed the listener. The bus drain becomes persistence-only -- the store is
the single fan-out point the UI process reads through its event pump.

The dashboard's configuration write path (PATCH /api/config, per-row reset,
per-repo field update) is deleted rather than left as dead policy: the
writes are descoped for the UI process (archie-core-ymut) and no route
reaches them. Editing is config.toml plus reload; the overlay store stays
for reload wiring.
2026-09-10 00:56:02 +10:00
sam b179ab92d0 feat(archieui): mount the capture intake receiver over the State Store
At the cutover (archie-core-8cda.5.4) the UI process is the only listener
serving POST /webhooks/capture/{source}: the receiver is a persistence shim
over the State Store contract, and the daemon's binding-dispatch loop keeps
consuming captures from the same store, so the HTTP front door moves without
moving the owner of work intake. Zero [capture] options fall back to
configuration.DefaultCapture; the arrival announcement persists through
InsertEvent so the event pump delivers it, where the daemon's bus drain
persisted it before.
2026-09-10 00:50:28 +10:00