jbergner
a6bc71fb3a
release-tag / Resolve release metadata (push) Successful in 30s
release-tag / Build knowledge (push) Failing after 4m51s
release-tag / Build control (push) Failing after 5m0s
release-tag / Build agent (push) Failing after 5m0s
release-tag / Build agent-data-init (push) Failing after 5m5s
release-tag / Build neuroforge-worker (push) Failing after 5m7s
release-tag / Build neuroforge (push) Failing after 5m9s
38 lines
2.2 KiB
Plaintext
38 lines
2.2 KiB
Plaintext
NeuroForge v0.8.2 validation
|
|
Date: 2026-08-25
|
|
|
|
PASS go test ./...
|
|
PASS go vet ./...
|
|
PASS go test -race ./internal/store ./internal/brain ./internal/httpapi ./internal/research
|
|
PASS Admin dashboard JavaScript syntax (node --check)
|
|
PASS OpenAPI YAML parses; info.version=0.8.2; 71 paths
|
|
PASS OpenAPI live research endpoints present:
|
|
GET /api/v1/goals/{id}/research/live
|
|
GET /api/v1/goals/{id}/research/history
|
|
PASS static linux/amd64 builds: server, worker, benchmark
|
|
PASS server smoke: /livez, /readyz, /version=0.8.2
|
|
PASS authenticated empty live-research endpoint response
|
|
PASS graceful SIGTERM shutdown
|
|
|
|
New v0.8.2 regression coverage:
|
|
PASS per-goal research run receives a persistent run_id
|
|
PASS research planning emits query.planned events
|
|
PASS SearXNG result discovery emits search.result events
|
|
PASS ingestion emits claim.extracted and evidence.learned events
|
|
PASS research run aggregates result/claim/evidence counters
|
|
PASS completed research run is linked from LearningCycle.research_run_id
|
|
PASS live endpoint returns only events newer than the requested sequence number
|
|
PASS live endpoint can reset when the latest run changes
|
|
PASS bounded ResearchRun/ResearchEvent storage prevents unbounded trace growth
|
|
PASS unfinished in-memory run is marked interrupted after restart
|
|
PASS source/memory durability remains on existing authoritative WAL/segment paths
|
|
|
|
Operational design notes:
|
|
- The dashboard uses authenticated incremental polling (~1.2 s), not SSE/WebSocket.
|
|
- Polling requests only the event delta after the last sequence number; it does not rescan all memories.
|
|
- Research trace UI events are accumulated in memory during a run and the bounded run is persisted at run completion. This deliberately avoids one fsync/WAL write per URL/chunk/UI event.
|
|
- Authoritative sources and learned memories continue to use the normal durable source/memory persistence paths.
|
|
- Claim entries shown live are transparent source-derived claim candidates/excerpts. They are not automatically labeled as verified truth.
|
|
- Download/source, duplicate/corroboration, rejection, and error lanes are represented separately in the UI.
|
|
PASS clean source ZIP rebuild: go test ./..., go vet ./..., server/worker/bench build
|