9 Commits
Author SHA1 Message Date
kyle 8b26776cdd feat(search): P3.1 filters and multi-query merge
Native include/exclude domains, time_range, topic, exact_match, min_score,
and up to 4 queries fused with RRF. Honesty notes stay off the Warn/fail
path. Tavily keeps include_answer false and sets auto_parameters false.
Cache key is canonical JSON. CLI flags and skill recipe included.
2026-09-29 15:30:19 -07:00
kyleandClaude Opus 5.5 a452aa16a1 P6 review: runner shutdown, cancel-while-busy, concurrent submits, batch-off 503, transactional store, CLI 202
Owner review of p6 (Ornith tasks 01-09), with an independent reviewer pass whose findings I
confirmed against the code. Every fix has a regression test that fails on the old code.

internal/batch
- Close let the workers keep draining every live job under the cancelled context, failing
  every target; they now stop at once, store nothing on a failure during shutdown, and leave
  the job live for MarkInFlightFailed("restart").
- A job cancelled while its in-flight target came back busy stayed in the rotation forever
  (depth gauge stuck, job held in memory); it now leaves. The task 03 design had this gap: mine.
- Submit's idempotency lookup, caps and create ran unlocked: 20 concurrent submits by one
  caller made 8 jobs. Serialised with submitMu.
- A failed target insert left a queued job no worker would run (holding the caller's cap);
  it is marked failed.
- A second concurrent cancel returned true and counted twice.
- Config: zero OR negative values take the defaults (negative Workers panicked the
  WaitGroup; a negative TTL or byte budget would sweep or evict everything).
- Retry timing used the pre-fetch time (a 5 s busy answer made RetryWait moot).
- The rotation cursor is kept within [0, len(live)] (it was absolute and drifted on removal).
- The unused in-memory callers map is gone (Submit racing Close wrote to a nil map).
internal/dcache
- PutBatchTargets, SweepBatch and eviction are transactional; the sweep never removes a
  queued or running job.
internal/api
- Every /v1/batch endpoint answers 503 when batch is off; POST panicked on the nil runner.
- GET /{id}/{idx} bounds idx by the rows actually returned.
- batchFetch copies a *fetch.Error instead of mutating one the PDF pool may share.
internal/config
- ParseSize no longer overflows to a negative size.
cmd/sift
- sift batch treated 202 Accepted (submit) as an error and exited 1 with no message.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-26 23:27:47 -07:00
kyleandornith-1.5-35b-a3b 2529cb8e84 dcache: batch job and target tables with a sliding lease, sweep and byte budget
Add the batch job store to internal/dcache (storage only; no HTTP, no worker).
New internal/dcache/batch.go holds the BatchJob/BatchTarget types and all
accessors; Open gains the batch_job/batch_target tables and indices, and DB
gains BatchH/BatchM counters.

Sliding lease: writes (UpdateBatchTarget, SetBatchJobStatus, TouchBatchJob,
MarkInFlightFailed) bump the job lease to now; reads never write. BatchJobByIDem
returns the newest (caller, idem_key) job with lease+ttl>now and status
queued/running/done/partial. UpdateBatchTarget counts targets_done exactly once
per target entering done/failed and adds len(result) to nbytes. SweepBatch
removes by lease; EvictBatchOldest evicts whole non-live jobs oldest-lease-first;
MarkInFlightFailed reconciles in-flight jobs on restart. Batch nbytes are kept out
of Size() and the content evict union.

Co-Authored-By: ornith-1.5-35b-a3b <ornith-1.5-35b-a3b@llama.cpp.invalid>
2026-09-26 20:38:42 -07:00
kyleandornith-1.5-35b-a3b b93b7d5816 dcache: PDF document cache (per URL and caller class, whole-document eviction)
Add pdf_doc/pdf_page tables plus PDFEntry, PDFKey, PutPDF (one transaction,
old pages deleted first), GetPDF (page-ordered, updates last_hit) and TouchPDF
(304 revalidation moves only fetched_at). PDF documents count toward Size() and
are evicted whole by oldest use; PDFH/PDFM counters added. evict() now breaks
last_hit ties by rowid so same-second inserts evict in insertion order.

Co-Authored-By: ornith-1.5-35b-a3b <ornith-1.5-35b-a3b@llama.cpp.invalid>
2026-09-25 23:05:12 -07:00
kyle 4372673184 p2.7: prometheus /metrics
Default GET /metrics is Prometheus text; JSON stays at /metrics.json
and ?format=json. :8443 needs the metrics caller bearer; loopback
stays open. Restricted callers have no per-caller series and no
tavily:restricted gauge.
2026-09-23 21:49:38 -07:00
kyle d7873403ff p3: search API, pinned sources, cache, CLI
POST /v1/search and /v2/search fan in SearXNG, Brave LLM Context, and
Tavily from config-fixed endpoints. Public sources refuse private A
records; SearXNG is a loopback-only client, not the fetch guard.

Cache keys include caller class. Restricted callers search like the
fleet; logs omit query, hosts, URLs, and upstream errors. Degraded
results are not stored. Tavily sends search_depth=basic and drops
answer. Missing keys surface in /healthz.

Default-on sources stay off until the bake-off. Query harvest is not
in this commit.
2026-09-23 12:17:51 -07:00
kyle 3687636ed7 p2.6: partition dcache by caller class; pin restricted workspace
Digest, chunk-embed, and question-embed keys include fleet|restricted so
a fleet caller cannot observe a restricted cache hit. Digest.Run takes
the class from permit_restricted.

A permit_restricted caller whose keys.toml workspace is not
guardian-star is refused at Search (no Honcho request) and flagged at
serve start and keys reload.

docs: P3 restricted callers search like the fleet; cache/logs isolate.
2026-09-23 11:51:39 -07:00
kyle 767c12c1af p2.3: pin chunk 0, hybrid score, keep question on retry
Question-mode selection always keeps the lead chunk, then fills
the budget by 0.7 cosine + 0.3 term overlap. prompt_version 2.
too-large embed retry no longer halves the question (index 0).
2026-09-23 03:18:29 -07:00
kyle 2f9d3d97e6 p2: digest with in-sift FIFO, SQLite caches, Honcho context, recover
Dixie is capped (max_inflight 2, queue_depth 8, queue_wait 20s) and
never queued on llama-server. Streaming chat, thinking off, 45s
fallback to clean. Digest and chunk-embedding caches are
content-addressed SQLite (Honcho context is not cached). Recover
uses the guarded Fetcher. guardian-star is refused in code.
2026-09-22 23:54:49 -07:00