Commit graph

808 commits

Author SHA1 Message Date
Meta Vens
f2cfdcbeb6 huginn-mind: import::plan owns its foreign keys
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:18:35 +01:00
Meta Vens
d215a6fa03 huginn-mind: import::plan lays out a moved campaign's replay, and the move tests replay through it
The verb and the tests share one definition of a replay batch: grouped by source receipt
ordinal, the Moved dropped from its own batch, foreign referents listed, oversized or
unordered receipts refused. Deletes the hand-built replay filter in move_tests.rs.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:18:35 +01:00
Meta Vens
0d4cccc157 huginn-mind: a campaign cannot be moved out of this mind's own instance root
A campaign slugged like the mind shares a root with the instance, its stewardships, memories and thoughts, so its Moved froze and closed all of them. A8 refuses such a Moved (MoveOfInstanceRoot); an import of the same campaign is refused already, the instance's records being native writes under its root.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:12:13 +01:00
Meta Vens
905c293396 huginn-mind: replay detection compares provenance
Whoever wrote a batch first decided its attribution: a native writer of a batch already imported was told AlreadyAdmitted and the record kept naming the source, and an import of a batch another source landed read the same. A9 now treats a stored receipt as a replay only when its imported_from equals the candidate's; otherwise the batch meets its own identities at A10 (IdentityCollision). A source naming this mind in another case is another mind, as the instance grammar already rules, and is pinned by a test.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:12:07 +01:00
Meta Vens
1533755425 huginn-mind: an import carries only a campaign's history, under a root only imports have written
Import mode turns derivation off, so it is open to any document whose derived writes native admission would have made: an imported hand-off left the steward's stewardship in force beside it, an imported ruling answered a question the Answered of which was never written and let a native ruling answer it again. A8 now refuses an import (ImportOutsideCampaign) unless each document's root names a campaign this mind holds or the batch carries and no native write has touched that root. Instance-rooted records have no campaign and never qualify. Replays that build a campaign from imports alone are untouched; the move tests that built the receiver natively now build it by import.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:11:56 +01:00
Meta Vens
3114e011e4 huginn-mind: a Moved resolution lands alone under its root
A sibling written in the Moved's batch froze with the campaign and never reached the receiver, whose replay stops before the Moved. A8 now refuses any staged document, derived ones included, rooted at a campaign whose Moved is in the same batch (CampaignMoved), on the giver and on import alike.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 19:05:45 +01:00
Meta Vens
b65eaa1fa9 huginn-mind: a moved campaign's closer is the Moved alone; the external listing names its referent
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 18:07:27 +01:00
Meta Vens
c4d38bcea5 huginn view: the header names an imported document's source and a moved document's receiver
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 17:47:43 +01:00
Meta Vens
c76e37280d huginn-mind: a Moved resolution freezes its campaign's root and the receiver replays the history as an import naming its source
The giver: (Campaign, Moved) joins the matrix; a Moved to this mind refuses MoveToSelf and a withdrawn Moved MoveIsFinal; any staged document rooted at a moved campaign, derived ones included, refuses CampaignMoved; Docs::closing_resolution reads every other document under the root closed by the Moved, and the image fans status_changed out so an earlier as_of still reads in force. The oracle derives the same fallback independently.

The receiver: PipelineProvenance.imported_from (optional, absent from every other receipt and from the digest) names the source receipt and the externals the batch cites. An import derives nothing, refuses its own instance, a Moved outcome, and an external this mind holds; a listed external stands in for an absent referent. mind-body persona.rs gains one outcome_words arm for the new variant.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 17:47:43 +01:00
Meta Vens
f29e15481e eureka-pipeline: a resolution outcome Moved { to } closes a campaign for the mind that gives it up
Additive resolution v2 variant, bounded as a Slug; the golden gains one line, the published schema one arm. Admission owns when it may be written.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 17:47:32 +01:00
91279f6496 Merge pull request 'mind-body: a nightly memory consolidation run for one Mind (typed Slept cut line)' (#14) from hands/eureka-body-sleep-turn-one-mind into main 2026-11-08 14:03:34 +00:00
Meta Vens
6aa5e01292 mind-body: the newest slept night's closing decides the day, not an older one's
A test with two slept nights pins that covered_through reads each night's own closing (the citation of its run), which no earlier test distinguished from the oldest Slept closing in force.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 14:24:07 +01:00
Meta Vens
7c9ccb9499 mind-body: the Sleep run is the night's memory consolidation only
SLEEP_GUIDANCE opens on 'Tonight you consolidate your memory.' and drops the contemplation paragraph; the night line closes on what she did with her memory, in the guidance and the schema's description. The module doc names the run the night's consolidation (rulings sleep-day-coverage-ruled, contemplation-own-session). A test pins that the start line carries the consolidation opening and that neither the guidance nor the schema invites her to think, read or look into anything.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 14:12:06 +01:00
Meta Vens
2e98327ded eureka-pipeline, huginn-mind, mind-body: a night's cut line is its typed Slept closing
ResolutionOutcome::Slept { line, day_through } is how a Sleep run ends when the ledger took her reply; day_through is the last memory of her day she was shown. CitationRole::DayThrough (day_through) makes the memory an A7 referent and a cites/cited edge; admission refuses Slept on any subject but a Sleep run, a day_through that is not a memory admitted before the batch, and skips the in-force check for that role since the night may have consumed it. The outcome alias carries Slept. sleep.rs close admits the typed outcome and covered_through reads only the Slept closing in force and day_through's admission ordinal: SLEPT, THROUGH and the rationale parse are gone, and a hand closing or a withdrawn Slept covers nothing. outcome_words renders Slept as 'slept: <line>'. Published schemas and the envelope golden regenerated from the derivation.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 14:07:22 +01:00
Meta Vens
ee5a4f92bb mind-body: a failed night after a slept one leaves the slept night's cover standing 2026-11-08 12:54:03 +01:00
Meta Vens
a621d03783 mind-body: public_surface names Night::from_settings 2026-11-08 12:32:32 +01:00
Meta Vens
2291e90355 mind-body: the once-per-window test frees her unit and moves the clock, so only last_sleep spends the window 2026-11-08 12:13:57 +01:00
Meta Vens
9a0dfdd0e1 mind-body: sleep day tests follow one-night-at-a-time admission and second-grained times 2026-11-08 12:08:05 +01:00
Meta Vens
9981c5b902 mind-body: read a closing's reason as its line 2026-11-08 12:04:16 +01:00
Meta Vens
e3be529b1e mind-body: a night's day is anchored on the newest night that slept; the heartbeat's night rules are pinned
The day began at the previous Sleep run whatever became of it, so the memories past DAY_MAX and a failed
night's whole day never met the disposition cover-the-day requires. A night now closes with where its day was
covered through (the ordinal of the last memory it was shown), and the next day is what was admitted after the
newest night that closed 'slept:'; the first night's day is the 24 hours before it opened. A consolidation is
made by a night, whatever room it keeps, so it is not the next day's memory.

Night::from_settings carries the both-or-neither start check out of the binary where a test reaches it.
Tests pin last_sleep with the unit gone, the window's minutes and its 23 and 59 bounds, and the half-set start.
2026-11-08 12:03:18 +01:00
Meta Vens
cebac9d426 mind-body: a Mind sleeps once a night, in her Persona slot, disposing of her day by typed operations
The spawn prelude leaves one_process for a shared start_session. sleep.rs is the one owner of a night: the night line (her day whole, open items, the tide), SleepReply and its schema, the decoder's ledger (every day memory exactly once, ids her memories in force, one conversation per consolidation, told back, third refusal ends it), admission of whole operations in batches that never split a consolidation, and the closing of the run with her line. The heartbeat opens one Sleep run per listed idle Mind per window (HEARTBEAT_SLEEP_UTC, HEARTBEAT_SLEEP_MINDS); a missed night is not made up and a failed one not retried.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 11:01:39 +01:00
Meta Vens
79edda37f6 eureka-pipeline, huginn-mind: a Sleep turn shares the Persona slot, a consolidation is a pair
RunTurn gains Sleep and RunSlot/RunTurn::slot (PersonaTurn and Sleep hold the Persona slot, a Self run the other); live_holder, launch holders and the unit template compare slots, not turns. MemoryOrigin gains Sleep { run }. Admission refuses an unpaired consolidation both ways (MindRefusal::UnpairedConsolidation): a memory that consolidates others without a Superseded resolution of each by it in the batch, and a memory Superseded by memories that do not name it. Published schemas regenerated.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 11:01:39 +01:00
8c85ee4455 Merge pull request 'chat_image: test that fails if the take(MAX+1) read bound is removed' (#12) from eureka/chat-image-read-bound into main 2026-11-08 09:36:07 +00:00
32bc2367da chat_image: the bounded read is its own function, with a test that fails without the take bound
load() refuses an over-cap file on its length first, so the take(MAX_BYTES + 1) on the read was
reachable only by a file that grew after that check and no test could see it. read_bounded holds the
read; its test grows a file past the cap and requires the read to stop at MAX_BYTES + 1.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-08 08:13:35 +00:00
c1c3e7d6fe Merge pull request 'mind-body: chat images reach the Persona session as image blocks (ChatDelivery v2 image field)' (#11) from eureka/chat-image-consumer into main 2026-11-08 08:12:44 +00:00
cd15cae22c persona tests: the whole-line golden carries the image trust sentence
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 20:55:37 +00:00
4aeaf2d6e7 persona: an image's content is data, never instruction, in the messages header
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 20:54:57 +00:00
e2c9638f27 mind-body: a delivered chat image reaches her session as an image block
ChatDelivery gains tuple index 11, `image: Option<ChatImage>` ([path, mime,
size, sha256], a reference into her per-room image store, never bytes); the
pending kind is gamecult.persona.chat_pending.v2, and v1 answers and records
still decode with no image. chat_image::load is the one place a reference
becomes bytes: exact `<hex sha256 of room id>/<sha256>[.ext]` path with no
traversal or symlink, regular file, 5 MiB declared and actual, declared size
and sha256, png/jpeg/gif/webp magic bytes matching the declared mime. The
persona line carries the image as a base64 image block after its text and an
`attachment role=organ` line saying attached or not shown and why. The
stream-json echo of a line with blocks still counts as our echo. The store is
MIND_PERSONA_IMAGE_STORE; unset, every image is not shown.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 20:52:30 +00:00
87b4f627b1 Merge pull request 'huginn admit holds near-duplicates; huginn recall dispatch' (#10) from eureka-substrate/huginn-recall-dispatch into main 2026-11-07 12:07:42 +00:00
Meta Vens
c8b338598c recall cited(): skip non-standing positions before scanning the word run, so one long run costs O(n) not O(n^2); a 1 MiB run test
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 12:58:03 +01:00
Meta Vens
337521f654 recall dispatch tests: box the held view 2026-11-07 12:32:21 +01:00
Meta Vens
9295b4c348 recall dispatch: cited scan without open loops; tie and dotted-slug tests 2026-11-07 12:08:42 +01:00
Meta Vens
28412f0da6 recall dispatch: d2 top-k test 2026-11-07 11:54:23 +01:00
Meta Vens
a34d54985b recall dispatch: block test budgets 2026-11-07 11:52:08 +01:00
Meta Vens
07b2d124e7 recall dispatch: fix literals, boxed view 2026-11-07 11:51:04 +01:00
Meta Vens
5b469ef54d huginn recall dispatch: the Agent PreToolUse hook appends the in-force neighbours of a brief's cited records; admit's hold check bounded at 10 s
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 11:49:48 +01:00
Meta Vens
373d96691f huginn admit: hold a batch once when its rulings or questions have in-force neighbours
cut huginn-admit-hold r2. admit decodes, then asks near::neighbours (the one
neighbour query, for admit now and recall dispatch next) for the nearest
in-force rulings and questions to each of the batch's first three titled
rulings or questions, minus the batch's own ids, the question a ruling answers
and the subject of a Superseded resolution. A neighbour left holds the batch:
exit 6, the neighbours and a token (16 hex of sha256 over the decoded batch and
the agent) printed, nothing sent. The same command with --seen <token> admits
it; any other token is InvalidInput. A trouble during the check prints a note on
stderr and the admission proceeds.

PipelineDocumentSummary::of is public so the neighbours print as header lines.
sha2 joins eureka-state (already in the lock).

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 11:35:01 +01:00
6c9581cb18 Merge pull request 'huginn persona get|put' (#9) from eureka-substrate/huginn-cli-json-persona into main 2026-11-07 10:06:28 +00:00
Meta Vens
bea5e8eccc huginn CLI: persona get and persona put
persona get prints the stored gamecult.persona_state.v0 document, then its
updatedAt and receipt (absent, exit 1, until the first put). persona put reads
the whole document from a file or stdin and prints Committed <receipt>, or the
refusal (a stale put names the stored updatedAt), exit 1; a document that is
not one JSON object is a usage error and sends nothing. Provenance tool is huginn.
The eureka-state MCP's last unique capability is now in the CLI.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 10:43:36 +01:00
c63327ece9 Merge pull request 'Release: Self replies reach her Persona; one conversation at a time; resume; project-name addressing; stream thoughts' (#8) from eureka-body/release-reply-path into main 2026-11-07 09:09:21 +00:00
Meta Vens
77ce0465ea stream: the proxy-environment test is an integration target, so the lib carries no environment read
Moves the_stream_agent_ignores_the_proxy_environment to crates/mind-body/tests/stream_proxy.rs, as control.rs carries its env-touching tests. It asserts through the public surface: with a proxy in the child's environment the stream's agent reaches its endpoint directly and follows no redirect, and a plain ureq agent in the same environment takes the proxy. No behaviour change.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 09:37:17 +01:00
Meta Vens
98a3741d4b stream: the binary's argument reader carries the reasoned allow its siblings carry
mind-stream reads its own arguments like the other binaries, which each allow the crate's disallowed-methods lint with a reason; one range assertion written as contains. No behaviour change.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 08:11:16 +00:00
Meta Vens
b90d84cd50 release-reply-path: two fixtures reconciled after the merges
A DM's room is its room id under the reply cut (the steward cut's hand-built DM delivery carried the alias form); a request after a result opens a turn under the task cut, so the resume cut's bound test closes that turn with a result.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 08:09:24 +00:00
Meta Vens
c1f7d7d8d9 Merge hands/eureka-body-stream-thoughts (e2e0488) into eureka-body/release-reply-path
# Conflicts:
#	crates/mind-body/src/persona.rs
2026-11-07 08:03:05 +00:00
Meta Vens
d920e8b171 Merge hands/eureka-body-project-name-addresses-steward (25502a5) into eureka-body/release-reply-path
# Conflicts:
#	crates/mind-body/src/persona.rs
#	crates/mind-body/src/persona/tests.rs
2026-11-07 08:02:11 +00:00
Meta Vens
8616cc2752 Merge hands/eureka-body-self-session-resume (46800b8) into eureka-body/release-reply-path
# Conflicts:
#	crates/mind-body/src/self_run.rs
#	crates/mind-body/src/self_run/decide.rs
#	crates/mind-body/src/self_run/decide/tests.rs
#	crates/mind-body/tests/public_surface.txt
2026-11-07 07:57:32 +00:00
Meta Vens
c5a2c68cb8 self: silent-stop kinds cover every kind the CLI queues nothing for on our stop; capture fixtures killshell and deadline-race; tests kill hand mutants H2 and H5
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 07:44:59 +00:00
Meta Vens
21f92068fb self: public_surface snapshot carries Usage's kind fields
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 07:55:40 +01:00
Meta Vens
7c0313b068 self: a stop of ours owes no turn only for a shell task (task_type local_bash, Bash or Monitor); every other or unread kind owes its turn when stopped
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 07:48:13 +01:00
Meta Vens
62ce2ba12d self: idleness reads the CLI's own background-task level and its notification-to-turn rule; at idle the oldest waiting conversation goes first
Cut: State.tasks, Event::TaskStarted, the tool_use_id keying of task lines and the notification replay, and TaskNotified's tasks.remove gate. running is a copy of the CLI's last background_tasks_changed, replaced whole; a notification owes a turn when the task was backgrounded and the session did not stop it with TaskStop (or KillShell, KillBash: the stream shows the alias as the model named it); Polled and Inbox carry other_first, so a served conversation's newer brief no longer outruns an older brief of another. Fixtures bg-fgto, bg-kill and bg-monitor pin the three falsified cases; the probe's fake API gains a generic tool step and a timed scenario.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-07 06:48:03 +01:00