Repository Issue Activity (beta)

#1833 [Architecture Proposal] Hardware-Gated Containment Framework (Genesis Protocol V5.0) for Recursive Self-Improvement

#1832 Proposal: harness-level failure attribution in eval reports (model vs harness vs eval definition)

#1825 Proposal: detect tool-trajectory regressions before deployment

#1829 Proposal: RES actor-indexing behavioral eval (20 samples, no custom code)

#1826 False-positive "distillation" ban on $200/mo Pro subscriber — automated appeal rejected by bot, requesting human review from Trust & Safety

#1827 Proposal: deterministic eval for agent action-boundary violations

#1824 Proposal: Shared Sparse Latent Neural Bus for Local ↔ Cloud Models

#1819 HttpRecorder silently discards logged eval events; HTTP error responses never trigger the fallback at any failure rate

#1818 API retry helper retries forever (no max_tries/max_time) and the error-detection branch in its callers is dead code

#1712 Model-graded classify: unparseable judge output silently becomes minimum score

#1817 [Bug] Context/Token Explosion in Geometric Script Generation (CadQuery/CalculiX)

#1811 LangChain completion wrappers drop runtime kwargs

#1810 Steganography does not flag missing required protocol fields

#1809 SchellingPoint mutates process-wide random state

#1805 MakeMeSay get_content fails on its documented dict response type

#1804 ML Agent Bench crashes when an action omits Action Input

#1803 Ballots role reversal rewrites system instructions as assistant messages

#1799 Post-hoc control recall counts undefined no-control samples after parse failures

#1798 Empty prompt lists are misclassified as chat prompts

#1797 LangChain math adapter can corrupt answers while removing Answer prefix

#1793 Tool conversation runner overcounts exhausted runs by one turn

#1792 TrackTheStat caps a perfect 300-turn run at 299

#1791 BluffEval can leave registry logging disabled after failed solver lookup

#1787 FunctionDeduction parser merges comma-separated integers without spaces

#1786 MakeMePay reports success fractions as percentages