Review fix for PR #873. The previous attempt wired Router.Restart on the
Telegram router, but that path is unreachable: telegram.go:260 registers an
exact-match /restart handler (restartHandler, restart.go:24) which consumes the
update before the chat router is reached.
The wiring was therefore a false affordance: it made the Telegram router report
RestartAvailable: true for a callback no production path invokes, and the test
only passed because it assembled a LocalChatAdapter by hand rather than driving
the composition production uses.
Remove it. Telegram's own restartHandler remains the sole path to
RequestRestart. The removals of boot.restartTelegram and
telegramSetup.RegisterRestart stand: those were written and never read, which is
the real defect (channels-telegram-3).
Tests are retargeted and table-driven: buildTelegramRouter leaves Restart nil
with and without an adapter (regression guard against re-adding the unreachable
wiring), the router's /restart route reports not-configured, and
RequestRestart's queue-then-reject contract is pinned at the adapter level.
This deliberately does NOT fix dashboard /restart: the web router runs in the
archie-gateway process while the Telegram adapter runs in archied, and no
gateway->daemon restart RPC exists. That needs its own design (see the bead
referenced in the PR body).
/restart could never work in production. boot.setupGateways captured the
Telegram adapter's restart callback into boot.restartTelegram, a field no code
ever read, so gateway.Router.Restart was nil in every composition:
handleRestartAdapter always answered "not configured" and LocalChatAdapter's
snapshot always reported RestartAvailable: false.
Fix at the producer: buildTelegramRouter assigns Router.Restart directly from
the adapter it already holds. The dead field and the telegramSetup.RegisterRestart
indirection are deleted rather than kept alive.
Red-then-green: TestBuildTelegramRouterWiresRestartCallback failed with
"router.Restart = nil, want the telegram adapter restart callback wired into the
router" before the change. TestBuildTelegramRouterRestartUnconfigured pins the
unchanged no-adapter behaviour, including the not-configured reply.
The selection criteria alone left workflow optional and defaulted an omitted
or unrecognised value to implement, so a decide call could still skip the
decision and send an unsettled capability straight to the builder.
Omission is now a retryable rejection: CaptureTool gains RequiredWhenTrue, a
conditional requirement a flat RequiredFields entry cannot express, because
the close-without-change branch has no workflow to name and must not be made
to invent one. A rejected call is not recorded, so it does not consume
MaxCalls and the model retries with the field.
A value outside the enum now fails the stage rather than degrading to
implement: reaching OnResult with one means the model ignored its own schema.
Splits validateCaptureArgs into per-constraint checks (cyclop) and carries
the regenerated contracts.json.
The classify stage picked between implement, tdd and feasibility with no
stated criteria: the mission spent its guidance on needs_code_change and the
schema said only 'Which workflow fits best', with implement as the free
default for an omitted or unrecognised answer. Issue 813 was classified
implement on that basis.
Both surfaces now state what each workflow is for and say not to pick
implement merely because nothing else obviously fits. The mission moves to a
named triageMission so a test can assert the criteria reach the model.
Closes archie-core-s5ys
labelKinds is keyed on bare strings, and this repo namespaces its labels
(type::feature, priority::medium), so KindsForLabels matched nothing and
every issue fell through kind routing to Route's triage fallback. Kind
routing has therefore never fired here: issue 813, a type::feature that
defaultKindWorkflows binds to feasibility, reached implement instead.
Labels now resolve to the value after the last '::'. The namespace is not
checked: bug/feature/bootstrap is specific enough that a label ending in one
means that kind whatever the tracker files it under.
Closes archie-core-tjoc
Timer.Stop cannot unwind a max-uptime callback that has already begun, so
cancelling still let the reaper and Release both stop and remove the same
container, which is the duplicate teardown and the misleading 'No such
container' warnings this was meant to remove.
Teardown is now claimed once per container: whichever of the two gets there
first talks to Docker and the loser skips to its own bookkeeping. Release
still claims after the grace period, so MaxUptime keeps biting during it.
Also table-drives the event pump priming cases, per the repo's test directive.
The reaper was cancelled at the top of Release, before the grace-period
sleep, which let a container outlive MaxUptime by GracePeriod. MaxUptime is
documented as a cap from creation enforced regardless of task state, and a
grace period is task state, so the cancel belongs after the sleep.
Also asserts the disarmed reaper directly rather than only inferring it from
the absence of later Docker calls.
The timer stayed armed after Release removed the container, so an hour later
it stopped and removed an ID Docker had already forgotten and logged two
warnings that read like a failed teardown.
archie-ui binds before archie-state-store is guaranteed to be listening, so
the first prime is routinely refused. PumpEvents returned on that error, which
left the activity feed history-only for the life of the process. Prime now
retries with capped backoff and gives up only when the context ends.
Orders archie-ui.service after archie-state-store.service as well, which
narrows the window without closing it: After= sequences the start, it does
not wait for the listener.
printf '%s\\n' emitted the literal characters backslash-n, which
archie-update-check's release-version check rejects, so every managed-agent
update left the installed agent version reported as unknown.
internal/tools/builtin.RegisterBuiltins has always accepted a GrepIndex and
nothing ever passed one, so every grep ran the direct walk no matter what
[indexing] was set to. internal/indexing was fully implemented and tested
with zero production callers, and internal/config decoded IndexingConfig into
nothing. Two halves of one defect.
The composition root now builds an indexing.Manager and hands it to the
workspace tool provider, which forwards it to RegisterBuiltins. deadcode
confirms the result rather than the commit message asserting it: unreachable
functions drop from 88 to 73, and all 15 internal/indexing entries are gone.
Failure degrades rather than propagates. A manager that cannot build logs a
warning and grep keeps its direct walk, which is exactly what it did before
this was wired. An accelerator must never stop the daemon starting.
The context is detached with WithoutCancel rather than replaced with
Background: the index build's lifetime is the process, while the gateway
path's boot context is cancelled once startup returns, so inheriting
cancellation would kill the build mid-flight. Background would have satisfied
the linter too, but would silently drop the caller's context values, so
registerTools now takes the context its callers already had.
internal/agentexec's RegisterBuiltins call is deliberately left alone. It
registers builtins only to harvest their names for a plugin tool-name
collision check and never executes them, so there is no grep there to
accelerate.
The agent-side half of the same gap. A stage failure already logged
"stage failed -- parking" before parking, but a run whose OUTCOME was parked
(a review that could not run, a workflow that ended with no outcome) reached
the timeline as a KindParked event and the store as park_reason and never
the log an operator downloads.
park is the one place a park's reason is final, so it is now the one place
that logs it; the stage-failure branch no longer logs its own near-copy, and
the stage names itself inside the reason. The entry carries the stage as a
field so nothing is lost by dropping the duplicate line.
observability.md's "Implemented design" section documented the container's
system-log path as though it were the only source of task-log content; it
now records the daemon's and the workflow's own park writers, including why
the entry follows the transition rather than preceding it.
An operator who asks "why did task 1 attempt 3 park?" and downloads that
attempt's log got a zero-byte file, because nothing daemon-side wrote to a
task log at all: the container's system-log subscription was the only
writer, so an attempt that parked before any container produced output --
a worktree prepare failure, an unavailable worker, a taskrun error -- left
the file the sink had created empty and the reason only in the daemon-wide
log.
Both park paths now write the reason into that attempt's sink. The entry is
written after the guarded transition lands, so a park another actor won
does not leave a log claiming an outcome the task board contradicts -- the
same race the parked event is already withheld for.
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.
A container run only restored ownership of the bind-mounted worktree on the
CommitAll/Push success path (archie-core#520). The agent runs as root inside
the container, so every other way a run can end -- a gate failure that parks
the task, a stage error, an early return -- left the worktree owned by UID 0
on the host. The daemon is the process that later cleans and resets that
directory as its own non-root user, and it cannot even unlink inside a
root-owned directory, so the next attempt failed before it could start no
matter what the worktree refresh did.
runTask now defers the same reconcileOwnership the commit and push paths
already call, so it runs at every exit. The retry recoveries in worktree
package cannot fix this from the daemon side: a non-root process cannot
chown, and removing the directory needs write access to it.
Observed through a supplementary group -- a non-root process may move a file
it owns into a group it is a member of -- so the test proves the chown ran
without needing root.
Residual, filed as archie-core-zcg8: a container that is SIGKILLed never
runs this defer.
A retry after any attempt that did not commit failed with
worktree prepare failed: checkout branch <branch>:
a branch named "refs/heads/<branch>" already exists
on a branch that was never the problem. refresh inferred "the branch does
not exist yet" from a failed checkout and reported whatever the Create
fallback then returned, so the real failure was replaced by a false one --
and the operator had no way to tell that the park reason was fiction.
The real failure, reproduced: a stage running in the container can leave a
path whose TYPE contradicts the tree refresh resets onto -- a directory
where the base tree has a file, or a file where the base tree has a
directory. go-git's hard reset cannot write over that contradiction and
fails with `openat <path>: is a directory` / `not a directory`, which is
exactly the failure the Create fallback then relabelled. Retry could not
recover from it, and a second path of the same class (a branch removed
while HEAD still pointed at it) failed with "reference not found".
refresh now clears the worktree against the tree it is resetting ONTO
before it checks anything out, which removes exactly those conflicts -- an
on-disk directory the tree declares as a file is untracked, and so is a
file the tree declares as a directory -- and resolves the branch's
existence from the reference store instead of inferring it. Resume gets the
same treatment, so the remediate path cannot fail the same way.
This is the third fix to this function (Force in 05944b6f was the second),
so the checkout/reset/clean sequence is rewritten rather than patched again.
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".
The download sets its attachment headers before the read, because the body
streams straight from the reader and a log is unbounded input -- buffering it
to size the response would defeat the reason it streams. Nothing then cleared
those headers, so a 404 for an attempt with no log went out as
Content-Disposition: attachment; filename="task-1-attempt-0.log"
Content-Type: application/x-ndjson
A browser treats that as a download: the caller gets a .log file containing
the text "no log recorded for this attempt". A stale or direct download URL is
enough to trigger it. The 503 for a process with no reader had the same shape.
Both error paths now clear the attachment headers before writing, so the
response is an error rather than a file. The successful path is unchanged.
Red first: the existing test asserted the status only, which is how this
passed. It now asserts the headers are cleared on a 404 and on a 503.
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.
One task attempt's persisted log lives in the state directory the State Store
process owns, and no contract carried it: the dashboard process therefore had
no way to read a log at all, and answered every attempt with "task logging is
optional and was not enabled for this run" in a deployment where logging is
unconditional.
ReadTaskLog returns a decoded page and StreamTaskLogContent returns the raw
bytes server-streamed -- a log rotates at 20 MB, past gRPC's unary cap, the
same reason StreamCaptures superseded ListCaptures. found=false with no error
means the attempt has no log; a service with no reader answers Unavailable,
and keeping those apart is the point: only the second is a deployment
condition.
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.
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.
Two defects in the activity/answer budget, both found in review.
The tool block returned as soon as the newest line alone fit, so a turn with
several short tool calls discarded history that was never over budget and
reported a "+N earlier" that was untrue -- those lines had room. The window
now widens from the newest line backwards while the result still fits, so
every line that fits is kept. The indicator appears only when something is
genuinely left out.
The opening frame sliced the answer to its budget without an ellipsis, so a
large first chunk -- or a turn that never opened a reply, leaving messageID 0
-- rendered text that appeared to begin mid-sentence, while later frames
marked the same cut. Telegram messages are immutable, so it could never be
corrected. Every frame now marks the cut; the special case only ever bought
one rune.
TestLiveReplyFramesStayWithinOneMessage asserted the ellipsis on the last
frame only, which is what let the unmarked opening frame through. It now
checks every frame, and the count bound follows: identical frames are
deduplicated by render(), so an unmarked opening frame was previously a
distinct body where a marked one is not. Verified by mutation -- restoring an
unmarked opening cut fails the test on frame 0.
Rebased onto main, which now carries an equivalent STATE_STORE_TOKEN isolation
guard in state_store_client_test.go from an independent baseline repair. The
two changes were the same fix expressed twice, so main's wording is kept and
this branch's duplicate is dropped -- the file is now identical to main and no
longer appears in the diff.
That leaves one commit, touching only internal/channels/telegram/live.go.
Removing clampToOneMessage, which this change orphaned: all three of its call
sites were replaced with clampToRunes directly, and `unused` flagged it. Dead
code left behind by a rewrite does not ship.
TestShutdownWatchdogHangingPastLeashForcesExit gave the watchdog two leash
lengths (60ms) to force an exit, but the 30ms leash only starts the timer --
the firing goroutine must then be scheduled before the injected exit is
observable. Measured margin is ~29.7ms, so the test is a wall-clock race
against the Go scheduler and fails intermittently on a loaded machine, where
it parks real work behind a gate failure unrelated to the change under test.
The assertion is now the contract that actually matters and that scheduling
jitter cannot falsify: the exit must not happen before the leash, and it must
happen. Waiting for it uses a generous ceiling, and the new elapsed >= leash
check catches an early fire strictly -- something the old test could not do,
because it only bounded lateness.
The inverse test asserts the ABSENCE of a forced exit, which load cannot
falsify (a slow runner only delays the exit being checked for), so it keeps a
tight bound and the suite stays fast.
Verified by mutation: removing the forced exit fails the first test, and
firing before the leash now fails it on the new elapsed check.
Review found that a bare "showing 3/197 lines" inside ordinary output was
accepted as truncation metadata and replaced by a summary, hiding the real
result. Both built-in producers wrap the notice as "[truncated: …]"
(internal/tools/builtin/truncate.go), so the wrapper is now required rather
than tolerated: it is the only marker separating a notice the tool printed
from content that merely mentions a count.
Four findings from review, all valid:
1. Search detection is now gated on the tool name rather than applied to every
result. A compiler diagnostic ("file.go:42: message") has the same shape as
a search hit, so counting it reported a build error as "N matches" and hid
the diagnostic.
2. Only ripgrep match lines are counted. The hyphen context form
("path-N-text") and "--" group separators were being counted as matches, so
a search with context requested reported the match count plus every
surrounding context line.
3. The notice parser accepts shell's "showing last N/M lines" form as well as
the plain "showing N/M lines" one. Shell emits the "last" form, which failed
integer parsing and was not recognised as metadata -- so a truncated shell
result quoted the raw truncation notice as its preview, the exact defect this
change set out to remove. The summary now also states which end was kept
("97 lines total; showing last 66" for a tail truncation).
4. The renderer cases are consolidated into one table-driven test per the
development protocol in AGENTS.md, covering all four defects above plus
singular/plural, colons in paths, and module-cache noise.
Operator report: tool output in the Telegram channel was unreadable during
search-heavy turns -- a wall of fenced blocks each showing one arbitrary
match line ("/path/config.json-322- }") under a heading that said
only "grep -- done".
The renderer quoted the first non-module-cache line of a result and appended
a line count recomputed from its own truncated sample, which could not agree
with the total the tool had already reported: one search rendered the tool's
"194 more lines" beside our "… 3 more lines". Two different numbers for one
result set read as noise, and the quoted line carried no context at all.
- Describe, don't quote: a search is reported as "N matches", a truncated
result as "<total> lines total; showing <n>", and every other tool falls
back to its first informative line. Truncation notices are skipped, since
they are metadata rather than content.
- Never invent a count: the tool's own reported total is the only
trustworthy source, parsed from its truncation notice rather than
recomputed from the visible sample.
- Remove the now-dead compactToolPreview and countNonEmptyOrContentLines.
Also bump go-telegram/bot v1.25.0 -> v1.27.0, which fixes a streaming bug
that lands squarely on this path: rawRequest sent the multipart body through
an io.Pipe, which net/http cannot replay, so sendMessage/editMessageText
failed until the connection was dropped.
Remaining work from the report (collapsing repeated calls into one counted
entry, and a ring of recent activity lines with a capped frame) is tracked
in #837.
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.
DELETE: IssueForge.RepliesAfter and its Reply type had no production caller --
the waiting_human status exists and is reachable, but automatic detection of a
human reply never landed, so the intended poller was never built. Neither
implementation paginated either (both read one page of 50), so the method
would have been wrong if wired as-is.
The manual path is untouched: waiting_human remains reachable and an operator
requeues through the existing task-action flow. Removing the interface method
also removes seven test fakes that existed only to satisfy it.
No behavioural test accompanies this: a pure deletion has no assertion to
make fail, and the compiler enforces completeness across every implementation.
WIRE: internal/taskstate documents itself as the single source for the
persisted task status strings, but store.go re-declared them as bare SQL
literals. Reference the workflow status constants (aliases of taskstate)
so a rename of a status is a compile-time-visible migration instead of a
silent change in what ClaimNext/RecoverStale/OpenPRs/ClearTerminalTasks
match.
WIRE: Config.MaxUptime and Acquire documented a hard container lifetime cap
(PRD section 4), but the code only wrapped ContainerCreate/ContainerStart in
a MaxUptime deadline that was cancelled when Acquire returned, so no running
container was ever killed. Wire the documented cap with a per-container
timer that stops and removes the container once the cap elapses.
WIRE: the SSE transport's Stop contract already says it closes the
connection and waits for background readers, but the reconnect loop issued
its HTTP GET with the Start context, which Stop could not cancel, so Stop
blocked on readerWg.Wait() against an unreachable endpoint. Wire a
per-lifecycle cancel (mirroring the stdio transport's stopCh plus guarded
commit) so Stop interrupts the in-flight reconnect request.
Lane F deleted skill.Discover/skill.DiscoverPlugins/skill.Skill (no
production caller) in 952bf4d, 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.
DELETE: no document in docs/architecture/ or docs/prds/ specifies a
production consumer for forge.IssueForge.CreateIssue. It has no
production caller -- the remaining repo matches are the GitHub
CreateIssueReaction and Gitea CreateIssueComment SDK methods. Removing
the interface method and the GitHub/Gitea/Noop implementations; the
other fakes that still define CreateIssue are test-only and out of lane
ownership, and an interface with one fewer method keeps them satisfying.
RepliesAfter and forge.Reply are untouched (a parallel FLAG record).
DELETE: no document in docs/architecture/ or docs/prds/ specifies a
consumer for binding.Matcher.Overlaps, and it has no caller or test
caller. The store enforces the overlap invariant with its own SQL
(internal/store/bindings.go InsertBinding/UpdateBinding), which is out
of this lane's ownership and left untouched. The doc comment's
empty-Source rule is already covered by Binding.validateCommon's
"matcher.source is required" check, so nothing needs moving.
WIRE: read and edit each declared a DefaultToolTimeout deadline and then
discarded it, so the documented contract (Tools that exceed this are
cancelled) was a no-op on the live tool-executor path. Wire the deadline
through runWithContext, the same helper write already uses, so a hung
filesystem cannot block the tool executor.