Commit Graph
8 Commits
Author SHA1 Message Date
morluto 443c7aff77 fix(evaluation): require explicit REA server attribution 2026-10-14 09:47:57 +08:00
morluto 1447fd2f99 fix(evaluation): share completed transcript facts across graders 2026-10-13 13:12:12 +08:00
Ysopwnable 844fa52772 fix(evaluation): authenticate delivered Evidence citations
Authenticate completed successful REA Evidence records for citation accounting and shared fixture grading. Preserve genuine structured and complete text-only delivery, including authenticated inventory substeps.

Closes #1728.
2026-10-11 14:48:46 +08:00
Ysopwnable 2653b194b6 feat(skills)!: improve Codex onboarding and task recovery (#1613)
Guide task-centered recovery from observed client, registration, PATH and
provider blockers. Install the recovery reference, document exact JavaScript
module selectors, and require clarification before searching for an unnamed
app. Validate public setup in isolated Codex accounts and preserve failed
or unrecognized tool outcomes in the existing evaluator.

Refs: #1587

BREAKING CHANGE: Optional verify:agent reports now use schemaVersion 4.
Consumers must accept expectedFirstTool: null and targetClarificationPassed
for missing-target scenarios. Production CLI/MCP contracts are unchanged.
2026-10-11 03:13:56 +08:00
YouWishandN0zoM1z0 27a9ceade7 fix!: restore caching, preserve failures, and verify fixture claims (#1211)
Restore profile-bound Ghidra snapshot reuse, preserve direct-analysis results through cleanup failures, and grade configured agent fixture claims against returned evidence.

Register the compiled agent-evaluation fixture entry for static analysis.

Co-authored-by: YouWish <29498139+Y0UW1SH@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
2026-10-09 14:49:09 +08:00
morluto 0a4caf4a0a fix(evaluation): canonicalize Unicode argument keys (#980) 2026-10-07 23:19:49 +08:00
morluto 2a6eb41147 refactor(tools): return complete analysis results inline 2026-09-29 14:53:50 +08:00
morluto 513d9b1555 test: overhaul deterministic suite architecture 2026-07-29 22:00:38 +00:00