The branch and main diverged at 8faa0f6a: the release commit was cut on
main while the branch kept advancing with the park-classification wave.
Merging main here first, so main can fast-forward to the branch tip
without a merge on the release side. No content conflicts: the release
commit touches the changelog and news artifacts only.
Domain fields ParkClass/RemediationRounds landed in d6f61c2c; this is
the wire and daemon half. Task gains park_class (32) and
remediation_rounds (33) and maps both directions; the classified park
write crosses the wire as ParkTask on a narrow TaskParker contract
(TaskLifecycle stays under the 8-method cap), with conformance
coverage local + grpc including the stale-guard sentinel and
needs_human normalization. Every daemon park site now states its
class: terminal for a repo that left config and a retired identity,
transient for the environmental parks (worktree, storage, grants,
taskrun), and identityMayAct returns the class beside its reason.
d6f61c2c added Task.ParkClass and Task.RemediationRounds to the domain
without the wire half; the parity guard went red. Add the two proto
fields (32/33) and map both directions in values.go.
Two beads, one root: four sites stamped a version and only two of them were CI.
`Taskfile.yml`'s build and `install.sh` derived GATEWAY_VERSION/RUNTIME_VERSION
from `git describe`, so a from-source build claimed a release it was not
running, and the stamp became a function of the commit graph -- the one input a
content-checksum build cache cannot see. An empty commit moved `git describe`
while `task build` reported "up to date" and shipped the previous stamp
(archie-core-pdq2). Both sites now stamp `dev`: a source build says so, and the
version a release carries is written into the artifact CI publishes. That
removes the non-file input rather than caching it, and traceability for a local
build is the checkout it came from.
`deploy.yml` published `:latest` from any push to main, so a release's own
stamping survived only until the next ordinary commit -- minutes, in an active
repo -- and a host deploying :latest reported dev, leaving the update path
unable to tell current from stale (archie-core-i5nb). `:latest` now moves only
from a commit that IS a release; every other build publishes `:edge` (mutable,
newest main) and `sha-<commit>` (immutable). A host that runs releases follows
`:latest`; a host that tracks main follows `:edge`, which RELEASING.md now says.
Verified: `task build` reports `archied dev`; the new workflow test fails
against the old workflow on all five guards and passes after; deploy.yml
parses; install.sh passes `bash -n`.
Gated with `task check` (exit 0) at 8faa0f6a -- the v1.39.0 base, CI-green --
with these five files applied. It is NOT gated at HEAD: HEAD carries an
unrelated red, `TestTaskProtoCarriesEveryDomainField`, because d6f61c2c added
ParkClass and RemediationRounds without the proto mapping. This change neither
causes nor fixes that.
The classification rides the API for free (taskView embeds task.Task),
so this is the read side only: a parked row shows what kind of
intervention it needs -- needs you, transient, terminal -- as a chip
with the answer in its tooltip, instead of a bare Parked every time.
Non-parked rows render nothing.
The F2+F3 schema wave from the auto-fix/resume investigation, sequenced
together as one migration. Every failure funneled to one parked status
with free-text ParkReason, and Task.RetryCount doubled as the
operator-retry budget and the review-remediation round budget -- so N
operator retries ate the remediation budget, a repo-level round cap
could re-park a task an operator had just legally retried, and no
surface could answer what a park needed from an operator.
Park class (F2): a producer-recorded vocabulary in taskstate --
needs_human (the default, so an unclassified park always reads as
operator-actionable), transient (environmental; a requeue can succeed),
terminal (retrying is pointless). Recorded at the park site, never
inferred from reason text; NormalizeParkClass makes unknown values
persist as needs_human. Transition-to-parked defaults the class, so
every existing unclassified park site reads correctly with no call-site
churn; the classified ParkTask write follows for parks that know
better. Requeue, RetryTask and BeginRemediation clear the class back to
the default with the reason.
Split budgets (F3): Task.RemediationRounds is the remediation's own
budget; RetryCount stays the operator's. The round-cap stage counts and
caps rounds; operator retry keeps counting RetryCount and can drive a
task to dead without touching rounds. remediation_rounds is a
presence-gated migrator column written through Update.
Known state: proto fields, the ParkTask RPC, the staterpc mapping and
the daemon-side park-site classification follow in the wire half of
this wave (the proto parity test is intentionally red until then).
Sequencing steps 6 and 7 of pr-review-remediation.md: the consumer
landed in 052e350b, and these are the two sources that feed it.
Webhook (decision 2, low latency): the GitHub receiver decodes
pull_request_review and pull_request_review_comment events behind the
existing HMAC verification and publishes the same typed reaction the
poller produces, keyed source-independently so PublishUnique collapses
the two sources. Only submitted/created carry a verdict to remediate;
edited and dismissed are acknowledged. Reaction deliveries are
rate-limited per source (decision 6's 'add when the receiver gains a
second event family'), and the reaction path carries no dispatch
predicate -- eligibility is the consumer's owned-PR lookup.
Poll (both forges, the no-webhook backstop): one shared scan over all
pr_open tasks, each with the forge client its own identity owns
(reconcilePRs's shape), listing reviews and comments since per-task
persisted high-water marks and skipping the owning identity's bot user
(decision 5's own-comment exclusion, which the poll is positioned to
make identity-correct).
The cursors are two, not one: review IDs and comment IDs are independent
forge sequences, and a combined high-water mark would silently skip
reviews once a larger comment ID landed. watch_comment_id -- dead since
the comment-watch design it was created for never shipped -- carries the
comment cursor; review_cursor is a new presence-gated migrator column,
and SetReviewCursors is its guarded RPC (pr_open only, so cursors never
move under a run that owns the task). forge.ReviewComment gains ReviewID
so a comment is collectable into its review's unit (decision 5) on both
forges, and forge.Review gains Body so poll-delivered reviews carry
their summary.
Note: the store/proto/facade half of this work landed early in a65eae24
under a release session's message -- a shared-checkout collision, kept
rather than rewritten; that commit carries the review_cursor migration
and the SetReviewCursors contract, this one carries the producers.
docs:artifact:check regenerated the artifact from a directory walk of
docs/, which also sees an UNTRACKED page. The committed artifact then
looked stale and the gate demanded a regeneration that was already
correct -- one that would have committed a page belonging to no commit.
It cost two sessions in one day: a migration branch, and an unrelated
writer adding a PRD.
Enumerate through `git ls-files --cached` instead, so a page the commit
carries is required in the artifact and a working-tree scratch file is
not. The index counts as well as HEAD, deliberately: a page staged for
this commit is about to be carried by it. Outside a git work tree there
is no commit to disagree with, so the walk is kept and says so on stderr
rather than passing as the stronger guarantee.
Test fails against the walk with the gate's own message ("is stale;
regenerate it by running ...") and passes against the enumeration.
Beads sync to a Dolt remote, so anything written into one is published.
Two beads filed earlier today named an operator host, its internal
address and its filesystem paths -- the tracker was treated as a private
store, which it is not.
The rule is stated where the tracker is described, next to the existing
one about .beads/ never reaching a commit, because it is the same class:
the tracker is not a private store. Host-specific facts belong in the
operator's memos instance, which a bead may point at.
The remediate workflow was registered but had no trigger: the reaction
stream was provisioned, the envelope, resolver and owned-PR lookup all
landed, yet nothing consumed the stream (pr-review-remediation.md's
sequencing left the consumer for this step). The daemon now drains the
reaction stream each cycle, before the task drain, so a reaction can be
claimed in the same pass that carried it.
Per decision 3, the store's owned-PR lookup is the authorization
boundary: an unmatched, merged or closed PR drops with a counter, and a
public reaction can never start an agent against a foreign repository.
Per decision 5, a review is the unit of remediation -- its comment
events are appended to the unit while it is queued and dropped once the
run is claimed (the round cap bounds the exchange); a standalone comment
is its own unit of one; approved reviews never remediate; and archie's
own comments, matched on the owning identity's bot user, are excluded so
its replies cannot retrigger it.
The guarded BeginRemediation/UpdateReviewPayload store methods are the
dedup: the pr_open -> queued transition carries the review unit in the
same write, so re-delivered reactions fail the status guard instead of
starting a second round. Both are new StateStoreService RPCs following
the Requeue shape, covered by the local/grpc conformance battery.
The maintainer approved task-run-master-detail.md after the
adversarial-review round; its status moves Draft -> Approved while the
work it describes is completed and verified.
A stage filter is a command scoped to the attempt it was made on: the
rail set it against that attempt's stage list. selectAttempt and
performRetry left filters untouched, so a stage selected on attempt 2
carried into attempt 1 -- a same-named stage there rendered as selected
without ever being commanded, or the log stayed narrowed to a stage the
new attempt does not have. The stage filter now resets when the selected
attempt changes; the level filter is attempt-agnostic and survives.
The master-detail PRD absorbs the adversarial review's verified findings:
the reload claim is narrowed to the URL contract (filters are not in the
URL), the unimplemented scroll and pinned claims are dropped, the
unnamed-stage and level-filter contracts are stated, the rail inventory
is corrected, and the acceptance list gains criteria for the
command-not-binding rule, attempt-switch reset, and keyboard selection.
The first cut grouped consecutive same-kind events, but the real noise
shape is an interleave: one background cycle emits curator_run,
session-memory and curator_action events carrying the same payload, and
kind-keyed grouping left them all visible. The grouping key is now the
rendered detail plus the task, so one cycle of noise collapses to one
row regardless of which internal kind emitted each event, while events
carrying different information never merge. Verified against a live
daemon: 50 background events collapse to one expandable row.
The five tabs gave the stage rail less than a third of the screen and
disconnected the log stream from the stage that produced it. The stage
rail is now the page's master pane, always visible; the four context
panels are the inspector and own the tab bar. Selecting a stage commands
the inspector to the log tab with that stage's filter set; clicking the
selected stage clears it. Selection is a command, not a binding -- the
operator can change the filter afterwards.
No endpoint, RPC or contract changed. tab=stages still resolves, to the
log tab; every inspector tab and attempt deep-link survives a reload.
Below the shell's 900px frame breakpoint the rail stacks above the
inspector. Design of record: docs/prds/task-run-master-detail.md (Draft,
per-run panel contracts unchanged from task-run-detail.md).
The feed's 50 raw events were dominated by consecutive same-shape
background rows (session-memory cycles, runner calls), which buried the
milestones the dashboard exists to surface. Consecutive events sharing
one label and one task now collapse into a group row with an xN badge;
expanding it reveals its events in place. The grouping is consecutive,
not global, so interleaved kinds keep their order and a milestone is
never merged with an event that merely shares its kind.
The projection is a pure view over the store's raw stream (activity-group.ts,
table-tested); the store keeps every event. Group rows keep ActivityRow's
link contract, and the disclosure button is stop-propagated inside the
first cell so it never fights the row-level link.
Carries the stack facts agents must not fight (token file, ui/dist as a
generated LAW asset, feature layout, UI gates), the operational-tool
design rules, the WCAG AA bar, and the rule that external design
prototypes are specs to translate, never code to port.
The info token doubled as badge text on info-soft and as the log-level
label, and failed 4.5:1 there (4.29:1 dark over card, 4.45:1 over glass;
light theme 4.11:1 on canvas, 3.58:1 on its soft). Dark --info moves to
#B08BF8, light --info to #6F47A6, both clearing 4.5:1 on every surface
they land on. Light --fg-subtle moves to #675E6F for the wash-composited
canvas where the old value measured 4.44:1. Measured pairs recorded at
the token definitions.
TestDeployWorkflowOnlyPublishesRuntimeImageForRuntimeTag pinned the
per-component contract: the agent image builds only when an archie/v* tag is
at HEAD. One-stream releases create only a v* tag -- release.sh has no
archie/v* arm any more -- so that selector can never fire again, and the
agent image would never be rebuilt. `docker compose pull agent` would fetch
an ancient image forever.
The test's own subject is still right and is kept: the gateway build stays
unconditional and the runtime build stays guarded. The guard now names both
selectors, and the comment says why the legacy arm remains (history is not
re-tagged). Caught by CI, not by me: I changed the workflow and pushed
without re-running the gate on that tree.
The one-stream change updated the changelog gate, release.sh and the docs,
but missed the image job. Measured on the v1.37.1 run: `gate` resolved
1.37.1 from the `v*` tag and then warned "No release tags at HEAD", while
`build-and-push` read only `archied/v*`/`archie/v*`, found neither, and fell
back to `dev+<sha>` for both components. It also skipped the archie-agent
image entirely, because that step was gated on an `archie/v*` tag being
present -- so a one-stream release would have published a dev-stamped
gateway and never rebuilt the runtime.
Three changes:
- version resolution falls back to the one-stream `v*` tag, so both
components are stamped with the release version;
- the runtime image builds on a `v*` release as well, since one version
covers the release; a legacy gateway-only release still skips it, which
is what release_tag exists to distinguish;
- the "no release tags at HEAD" warning recognises `v*`, so a real release
stops being told it is not one.
Dockerfile.archied still copied CHANGELOG.archied.md and CHANGELOG.archie.md,
which the one-stream release deleted. The image build therefore failed at
the COPY stage, so no images and no GitHub Release have been published
since the stream merged -- v1.37.1 failed the same way, and this release
failed with it.
Copy the single CHANGELOG.md the streams merged into. Verified by building
the image: the build completes and /usr/share/archie/CHANGELOG.md is
present and readable in the result.
Moves event capture, mappings and bindings onto PocketBase-owned collections
in internal/infrastructure/edastore, and adds the read-only stateadmin
surface over the task database. Follow-ups included:
- the event feed cursor becomes a total order over a fixed-width timestamp
with an opaque string cursor, so it no longer depends on rowid
monotonicity;
- cron schedules get next-run arithmetic, so a cron job fires instead of
returning ErrScheduleUnsupported;
- binding and mapping edits announce on the activity stream, so a dashboard
edit is visible without a manual refresh;
- an unresolvable task identity fails closed rather than falling back to the
root forge's credential;
- tool_call events project into the tool_calls collection, which had a
schema and no writer.
Also: the branch is brought back to a passing task check, which it had not
been for the whole migration -- two pre-existing failures, an unrecorded
int64-to-string proto break and a stale generated docs artifact.
task check's fmt pass rewrites both, so the committed tree was not
fmt-canonical. Adopted verbatim, per the repository's rule that formatting
changes from the configured formatter are law.
Lane D. The collection existed in the event-capture schema with nothing
writing it; tool calls reached only the on-disk task log, readable as text
and never queryable.
The projection is a decorator on the State Store's task store, wired in
stateStoreDeps. InsertEvent persists the event first and only then, for a
tool_call event, writes one row. A projection failure is counted and logged
but never returned: event persistence is load-bearing, the transcript is
observability, and failing a task write because its shadow failed trades
the fact for the record of it. With no event-capture store the task store
passes through unwrapped.
The event adapts to the existing collection rather than the collection to
the event; args and duration_ms stay reserved. InsertToolCall stays silent
on the notify hook, matching the dispatch tables rather than the
operator-editable ones -- tool calls are high-frequency telemetry, not
dashboard activity.
The tool_calls collection existed in the event-capture schema but nothing
wrote it. Give it its write path:
- edastore gains InsertToolCall and the minimal TaskToolCalls read, writing
only the existing collection. No new index: the collection is an
append-only transcript and the SQLite predecessor had no unique index.
- internal/app/archied composes toolCallProjectingTaskStore, a decorator on
the task store that persists the event first, then -- only for tool_call
events -- writes one row. A projection failure is counted and logged but
never returned from InsertEvent: event persistence is load-bearing,
tool_calls is observability.
- stateStoreDeps wires the decorator around deps.Tasks when both stores are
open, and passes Tasks through untouched otherwise.
The event adapts to the existing schema, not the other way round: the tool
name comes from data, a succeeded call's detail is the row's result, a
failed call's detail its error; args and duration_ms stay reserved for a
producer that reports them.
Lane B. A task naming a non-empty identity this daemon no longer knows --
renamed, retired or deleted in the control plane -- used to fall back to the
root forge's credential. It now parks, and a NATS intake carrying one is
rejected before it is enqueued. An empty identity is the single-identity
root and still resolves to root.
Validation lives at the intake producer rather than in a foreign key:
tasks.identity uses '' as the valid root sentinel, so an FK would have to
admit '' as well as a real id, and turning on PRAGMA foreign_keys to make it
bite would also start enforcing the identity_aliases/identity_subjects FKs
that are decorative today -- wider than this change.
0e6c6fc6 added docs/prds/artifact-sender.md without regenerating the
artifact, so docs:artifact:check failed at that commit. Generated from a
clean worktree at 0e6c6fc6.
Implements docs/prds/artifact-sender.md (Approved). archie-core now sends
generated artifacts to the offloaded.dev collaborative editor, whose
POST /api/artifacts route, bearer service token, PocketBase-backed
versions and idempotency semantics are the receiving contract.
- [artifacts] config section: base_url + token SecretRef with a
token_env fallback naming WORKSPACE_INGEST_TOKEN. An empty base_url
disables the sender entirely; a base_url with an unresolvable token
disables it with a warning, because a missing credential disables a
capability rather than stopping the daemon.
- internal/infrastructure/artifactsync: one-method Publish seam in the
Courier posture, full source per publish (versions of record are
snapshots), 503 retried three times with backoff, 401/400 surfaced
immediately with the field reason, deterministic request ids
(artifact:pr-review:<owner>/<repo>#<n>).
- ReviewReport gains a pure markdown renderer in internal/domain/workflow:
summary, one subsection per finding with the failure scenario and
disposition when present, verified properties as a checked-clean list.
- The daemon's PR reviewer publishes after a completed review when the
sender is configured, records the editor deep link on the result, and
degrades a failed publish to a log line -- publishing never fails the
producing work.
The cursor change renamed the field to 'cursor' as well as widening its
type, which buf breaking flags as FIELD_SAME_NAME and FIELD_SAME_JSON_NAME
on top of the type change. The name buys nothing: the field is already
documented as an opaque cursor, and the number and name are the wire
identity. Adopt the gate's verdict and change only the type, keeping the
break confined to the one rule already recorded in proto/buf.yaml.
Lane A. EventsSince no longer relies on events.id being monotonic: it
orders by (at, id), where at is written in a fixed-width layout so
lexicographic order is chronological order, and id is only a
deterministic tie-break. The cursor crosses the wire as an opaque string
and rides the SSE id: field, so EventSource's own Last-Event-ID resume
keeps working untouched. A malformed or legacy cursor degrades to
"from the beginning" rather than failing the feed.
events.id keeps its int64 type and row-identity meaning; no
push/subscribe RPC is added.
EventsSince no longer depends on events.id being monotonic. The feed now
pages a total order over (at, id): at is a fixed-width, lexicographically
sortable UTC timestamp (storecontract.EventCursorLayout) and id is only a
deterministic tie-break for rows sharing one timestamp. The cursor is an
opaque string on the wire (EventsSinceRequest.cursor), carried verbatim as
the SSE id: field and echoed back via Last-Event-ID, so EventSource resume
needs no client change.
A new idempotent migration normalises pre-existing RFC3339Nano at values to
the fixed-width layout before finishTaskMigration, and idx_events_at backs
the new query. Legacy integer and garbage cursors degrade to "from the
beginning", preserving the old failed-ParseInt-yields-0 behaviour.
Amends the "Dashboard live event delivery" decision's cursor-cannot-skip
justification, which previously rested on commit-order id assignment.
Lane C. A successful binding or mapping write on the State Store's
PocketBase tables now emits one activity event on the same task-DB path
capture arrival already uses, so a dashboard edit is visible on the
activity stream without a manual refresh.
The event is an invalidation signal, not a payload: it carries the record
id and the action, and consumers refetch through the authenticated API. A
nil notifier is a silent no-op, so stores composed without one (tests,
tooling) are unchanged.
docs:artifact:check has failed on this branch since e95a77c2 deleted
ARCHITECTURE.md and moved its normative content into docs/architecture;
the generated dev-docs artifact was never regenerated, so task check could
not pass. Regenerated from a clean tree (no untracked docs present).
The event-capture contract moved its record ids from the SQLite rowid
(int64) to PocketBase's string record ids in place, on the same field
numbers: 16 fields across CapturedEvent, Mapping, Binding,
RecordDispatchRequest, EnqueueBindingTaskRequest, Task.binding_id and the
Id-bearing request/response messages. buf breaking has flagged them since
that commit landed, so task check has never passed on this branch.
Both halves of the State Store contract ship from this repository
together, so they cannot be versioned apart. Record the break narrowly:
only FIELD_SAME_TYPE is relaxed, and only for state/v1/state.proto. Every
other FILE rule still applies to that file and no other contract is
affected.
Each successful write on the State Store's PocketBase bindings/mappings
tables emits one event (binding_changed / mapping_changed) through a
nil-safe Notify func injected at composition, so a dashboard edit is
visible without a manual refresh -- the same task-DB event path capture
arrival uses. Nil notifier keeps every existing constructor silent.
tools/cron ANDs the two day fields where Vixie cron ORs them, so
0 0 1 * 1 means Monday the 1st, not the 1st or any Monday. State the
deviation in the ScheduleCron comment and pin it with a test.
The event-capture tables (captures, mappings, bindings, binding_dispatches,
playbook_dispatches) now live in PocketBase-owned collections and their ids
are opaque record ids rather than autoincrement integers. The task lifecycle
stays on SQLite: tasks/transitions/events carry FIFO claim ordering, id-cursor
pagination and the task-grant map, none of which gain anything from being
hand-editable. binding_dispatches.task_id is the single seam and stays an
integer addressing that store.
BREAKING: CapturedEvent.Id, Mapping.Id, Binding.Id and Binding.MappingId are
strings on the wire, as are the ids in the capture/mapping/binding requests.
RecordDispatchRequest.task_id is deliberately unchanged.
What the collections buy, beyond storage: a payload nobody has modelled yet is
inspectable, a mapping is editable at runtime, and a binding's relation to its
mapping is enforced. The old schema declared no FOREIGN KEY on that link, so a
dangling mapping_id was previously possible and is now refused.
Behaviour preserved deliberately, each recovered from the SQLite version
rather than reimplemented from memory:
- draft -> pending_approval -> armed, with approval only from
pending_approval and an edit always returning a binding for re-approval
- one binding per source, guarded on insert, update and approve
- an empty secret on update keeps the stored one, so an edit form that does
not echo it back cannot silently disarm a binding
- ErrMappingNotFound / ErrBindingNotFound on a missing delete target
- sentinels are re-exported from storecontract, never redefined: staterpc
matches on exact canonical message to rehydrate them, so a second
definition would break errors.Is across the wire invisibly
Two implementation notes worth keeping:
- at-most-once dispatch is now a unique index rather than INSERT OR IGNORE.
Only the unique-violation maps to ErrAlreadyDispatched; any other save
error is returned, because treating them all as "already recorded" would
turn the ledger into a silent at-least-once dispatcher.
- captures store received_at themselves at nanosecond precision in a
fixed-width layout. PocketBase autodates are millisecond and record ids
are random, so same-millisecond captures would otherwise order
arbitrarily; the old store got insertion order free from autoincrement.
- ListUndispatchedCaptures is raw SQL: the anti-join is a subquery and
PocketBase's filter DSL is not SQL. A read fires no hooks either way.
Adds tool_calls, which had no persistence at all: tool calls only ever reached
the on-disk task log, readable as text but never queryable or watchable.
Nothing writes it yet.
The binding secret cipher moves with the bindings it protects.
The schedules feature no longer runs on it: jobs live in the state store
as the schedules control-plane document, read and written through the
control-plane client (scheduleResourceStore), and the legacy jobs.json
migration into that document has run its course. What remained of
internal/infrastructure/cronstore was 1,600 lines of file store, OS-level
cross-process locking and schema-versioned JSON envelope -- plus the
schedule vocabulary held hostage inside it.
The vocabulary moves to internal/domain/scheduling, which is where
migration-decisions already said a schedules vocabulary belongs: a domain
vocabulary package must not import infrastructure. JobSpec, Target,
Payload, the kind constants, Schedule and Duration move unchanged -- same
wire shape, same JSON tags -- with the two sentinel errors callers
matched outside the package (ErrJobNotFound, ErrScheduleUnsupported) and
ErrInvalidSpec, which the schedule arithmetic itself raises. The store's
own sentinels die with the store.
The legacy migration path (migrateLegacySchedules, readLegacyJobs) goes
with it: its only dependency was cronstore.Open, and its work is done.
crondelivery keeps its consumer-owned contracts; the compile-time
assertions that pinned *cronstore.Store move to the production
implementation, scheduleResourceStore, in internal/app/archied. The
package's fixtures swap the real file store for an in-memory one with
the same Get/MarkRun/Due semantics, so the run-recording acceptance test
still asserts against real schedule arithmetic.
A future storage backend (PocketBase) implements the same vocabulary.
First slice of the persistence migration, taking the event-capture tables
rather than the task lifecycle. These five tables hold no production rows and
their feature (archie-core-t2db) is unfinished, so they migrate at zero data
risk and the remaining work lands natively instead of being written against
hand-rolled SQL and ported later. They are also the tables that actually want
what PocketBase gives: inspecting a payload nobody has modelled yet, mapping
fields against it, and editing a binding at runtime are the epic's acceptance
criteria, not incidental conveniences.
The task lifecycle deliberately stays on SQLite. tasks/transitions/events
carry FIFO claim ordering, id-cursor pagination and the task-grant map, none
of which gain anything from being hand-editable. binding_dispatches.task_id
is the single seam and stays a plain integer addressing that store.
Idempotency moves from INSERT OR IGNORE on a composite primary key to unique
indexes, so the at-most-once ledgers are enforced by the schema rather than by
each caller remembering the pattern. A save that fails for any other reason is
returned rather than swallowed: treating every error as "already recorded"
would turn a broken ledger into a silent at-least-once dispatcher.
Adds tool_calls, which had no persistence at all. Tool calls were only written
to the on-disk task log (internal/logging), readable as text but never
queryable, filterable or watchable.
Both collection lookups treated any error as "not present". For the users
collection that returned nil on a real failure (caught by nilerr); for the
views it was worse but unflagged, because creating a fresh view collection on
the strength of a transient lookup error is a schema change made for the wrong
reason. Only sql.ErrNoRows now means absent.
PocketBase creates a "users" auth collection on first bootstrap whose
createRule is open, so an unauthenticated caller could POST a record and write
a row into the State Store's database file. This surface authenticates
superusers only and has no notion of end users, so the collection is deleted
rather than given a tighter rule: one that is absent cannot be loosened again
by a later PocketBase default.
Also re-applies the modernc pin. It was applied while developing the previous
commit but a concurrent go.mod resolution reverted it before that commit was
taken, so the tree has been running 1.59.0/1.77.1 against a PocketBase built
and tested on 1.57.0/1.74.4, printing the mismatch warning on every start.
Note for future dependency work: go get -u will undo this pin, and
modernc.org/libc does not follow semver, which is why PocketBase checks.
The claim in the previous commit message that the pin landed there is wrong.
Serve an optional PocketBase surface over archie.db from the process that
already owns it, behind archie-state-store -admin-addr. It co-tenants on
<db_path>-tasks.sqlite via PocketBase's DBConnect hook rather than opening a
database of its own, so the single-owner ownership claim covers it and
PRAGMA user_version stays archie's.
Every archie table is exposed as a view collection, which is read-only by
construction. Views alone do not hold that line: PocketBase also ships
POST /api/sql, which executes write statements as well as reads, and
/api/backups, which would put a second snapshot and restore semantics beside
store.Backup's VACUUM INTO. Both routes are refused rather than left behind a
superuser check, because a superuser writing task state is still a second
writer. The gRPC contract remains the only writer.
apis.NewRouter binds the JSON API but not the dashboard (apis.Serve mounts the
embedded UI itself), so /_/ is mounted here and the caller keeps the listener.
The dependency pin and the Taskfile/README changes for this landed with the
preceding commit, which swept up the staged files.
The index built google/codesearch over the chat workspace on every daemon
start, for the process lifetime, through a child-process helper and a
SQLite metadata DB -- and grep consulted it only for searches confined to
the workspace root, which the analysed sessions showed was a small
fraction of searches. The compute was continuous; the benefit was not.
Removes internal/indexing, the workspace-codesearch helper subcommand,
the [indexing] config section, and the GrepIndex accelerator seam from
the grep tool, which returns to its direct walk. google/codesearch
leaves go.mod with it.
ARCHITECTURE.md duplicated the docs/architecture tree and had drifted badly
against it: "internal/ holds ~36" against 112 packages, two of the seven
binaries listed, a Forge table listing `RepliesAfter` (deleted from the tree),
and a stated invariant ("No daemon-internal cron") contradicted by the
scheduling engine that internal/app/archied/scheduling.go composes and
main.go starts.
It was also the sole home of three pieces of still-authoritative content, so
those move rather than disappear:
- plugin engine rule (strict), the memory engine family, and the
internal/plugin/architecture_test.go enforcement note ->
docs/architecture/plugins-and-extensions.md
- intake provenance: forge poll/webhook dedup through
TaskEnvelope.IdempotencyKey, binding-stamped chat/playbook enqueue, and the
single-identity webhook-intake rejection ->
docs/architecture/messaging-and-work-intake.md
Live references are repointed from CLAUDE.md (AGENTS.md is a symlink to it),
docs/architecture, docs/guides, seven PRDs, and four .agents skills. The
heading "## Plugin engine rule (strict)" is preserved in its new home so the
existing #plugin-engine-rule-strict citations keep resolving. Generated docs
artifacts regenerated.