Repository Issue Activity (beta)

#773 Re-derive data_access_level for academic-paper-reviewer and academic-paper under the dirtiest-input rule

#769 Register write-scope guard launcher degradations in degradation_registry.json

#771 Consolidate the duplicated markdown-stripping helpers in check_control_availability.py / check_data_flows.py

#794 Second slice of the markdown-grammar consolidation (#771 follow-up)

#790 QA report for academic-paper-reviewer: Grade C (62/100) — free recommendations available

#787 Run the Promotion Bakeoff for gpt-5.6-sol over the ChatGPT-subscription codex transport

#789 Sealed preregistration for future Promotion Bakeoffs (hash-commit, reveal after the fleet)

#785 Codex subscription transport: login-status attestation read from stdout only, but codex 0.147.0 emits it on stderr

#783 Cross-model docs: recommendation surfaces still bind to superseded GPT-5.5

#781 Directory listing: add OrcaRouter to THIRD_PARTY.md

#775 Add OrcaRouter as a named OpenAI-compatible cross-model provider

#761 [META] ISO 42001-spirit audit (2026-08-17): transparency / verifiability / feasibility track

#760 Solo-maintainer governance statement + SECURITY.md triage substance

#759 Lightweight risk register: risk → control → evidence status → residual gap

#755 Document CI workflow enforcement classes (blocking / advisory / administrative / post-release)

#756 data_access_level: verified_only on academic-pipeline contradicts the dirtiest-input rule

#758 DATA_FLOWS.md: single statement of what leaves the machine, what is cached, and for how long

#757 Per-channel control-availability matrix (plugin / skills / clone / Claude Science / Pi / no-Python)

#753 Align distribution-surface claims with evidence ceilings (plugin.json / marketplace.json / no-bypass prose / unvalidated numbers)

#754 Version-consistency lint blind spot: CITATION.cff and POSITIONING citation prose drifted to 3.14.0

#743 Design a bounded inquiry branch ledger with park, reopen, merge, and invalidation semantics

#745 Publish a stage capability/evidence matrix with enforceable claim ceilings

#742 Define research-family workflow profiles and progressive-disclosure usability gates

#740 Replace binary reviewer independence claims with typed run provenance

#739 Remove uncalibrated reviewer-score precision and generic score-to-decision mappings