Commit Graph
9 Commits
Author SHA1 Message Date
Marius CiepluchandClaude Opus 5 0c2b9ec685 feat(learning-harness): canary snippets and the second judge
Task 2.7. A canary is how a family says the vulnerability fired. Injection
plants a row in an in-memory database that only an unconstrained query returns,
file path plants a file under the scratch mount that only a traversal reaches,
and deserialization plants an object that only an unsafe load reconstructs. Each
prints its token on that path and no other, and the token is derived from the
claim, so one family's oracle can never answer for another and one claim's
canary can never answer for the next. A snippet that cannot reach the function
prints nothing, which is a suspicion rather than a proof.

Every template passes the existing snippet safety check before it can run, and
the candidate is always imported from the read-only source mount.

Nothing reaches proven on a machine's say-so alone. A second, independent
judgement reads the snippet and what the sandbox printed and either confirms or
drops the claim back to corroboration; a judge that cannot answer never promotes
anything. The judge is told the claim and the output, never the answer key.
Templates and oracle code live in the package, outside every directory a round
may write.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01T7XD7x5KonMdmGRkp63EQG
2026-09-06 10:06:57 +02:00
Marius CiepluchandClaude Fable 5.1 8f444051f2 refactor(corpus-seeded-mechanisms): apply review round 2 suggestions
- variant_search.search_tree skips a malformed store row with a
  variants_store_row_invalid degradation instead of aborting the scan;
  a hit on a demoted or flowless overlay record is never merged (test).
- seed.pair_lessons is the one shape derivation shared by the exporter
  and leave-one-out; labels pass line and sink together; pairs without a
  labeled function are reported as no_labeled_function.
- config parses [variants].max_mechanisms with the shared int helper.
- regress.candidate.hotspot_from_finding is the one zero-score hotspot
  helper for forced inventory and variant findings.
- design synced (SearchResult, hits_to_findings summaries, candidate.py).

Co-Authored-By: Claude Fable 5.1 <[email protected]>
Claude-Session: https://claude.ai/code/session_01T7XD7x5KonMdmGRkp63EQG
2026-09-05 14:55:00 +02:00
Marius Ciepluch 818a0b10d3 feat(three-stage-scan): report map and honor fail-on worth-fixing 2026-09-04 00:18:33 +02:00
Marius Ciepluch a916e2d173 feat(three-stage-scan): attach tool hunter to map hotspots 2026-09-04 00:04:36 +02:00
Marius Ciepluch 2efbbb8024 feat(three-stage-scan): run or skip deep regression with visible degradation 2026-09-03 23:43:15 +02:00
Marius Ciepluch 331b62f1c1 feat(three-stage-scan): map sandbox outcomes to worth-fixing verdicts 2026-09-03 22:59:01 +02:00
Marius Ciepluch 62c7990e49 feat(three-stage-scan): add language recipes with inconclusive fallback 2026-09-03 22:48:10 +02:00
Marius Ciepluch 7a584d4578 feat(three-stage-scan): select bounded regress candidates with sev-5 extras 2026-09-03 22:36:02 +02:00
Marius Ciepluch eb87ef85ec feat(three-stage-scan): reject unsafe regression snippets before execution 2026-09-03 22:12:45 +02:00