2026-03-07 - 2026-03-14
Overview
1 Release published by 1 user
Published
hermes/v0.1
22 Pull requests merged by 3 users
Merged
#63 [loop-cycle-5] Persist chat history in SQLite (#46)
Merged
#62 feat: add --session-id to timmy chat CLI
Merged
#61 [loop-cycle-4] Push event system coverage to ≥80% on all modules
Merged
#58 [loop-cycle-1] Fix XSS vulnerability in swarm_live.html (#47)
Merged
#44 feat: Add Kimi agent workspace with development scaffolding
Merged
#43 fix: fact distillation — block garbage and secrets, improve dedup
Merged
#42 fix: persistent event loop in CLI interview — no more Event loop is closed
Merged
#41 fix: corrupted memory state + regex bug in update_user_profile
Merged
#33 cleanup: delete ~8,000 lines of dead code + sovereignty fix
Merged
#21 refactor: YAML-driven agent config — kill hardcoded personas
Merged
#20 fix: MCP integration — StdioServerParameters + smoke-tested
Merged
#14 feat: replace custom Gitea with MCP servers
Merged
#13 claude/sharp-mcnulty
Merged
#9 feat: add Gitea issue creation — Timmy's self-improvement channel
Merged
#8 feat: consolidate memory into unified memory.db with 4-type model
Merged
#7 feat: add DB Explorer for SQLite inspection
Merged
#6 feat: add Loop QA self-testing framework
Merged
#5 feat: enrich thinking engine — anti-loop, anti-confabulation, grounding
Merged
#4 feat: tick prompt arg + fix name extraction
Merged
#3 fix: test DB isolation, Discord recovery, and over-mocked tests
Merged
#2 fix: WebSocket crash from websockets 16.0 + branch pruning
Merged
#1 feat: upgrade to qwen3.5, self-hosted Gitea CI, optimize Docker image
3 Pull requests proposed by 2 users
Proposed
#57 fix: replace bare except clauses with specific types (#25)
Proposed
#59 fix: replace print() with proper logging (#29, #51)
Proposed
#60 [loop-cycle-3] fix: replace eval() with AST-walking safe evaluator (#52)
20 Issues closed from 3 users
Closed
#38 Thinking engine: break the rumination loop
Closed
#46 [HIGH] Chat history is in-memory only — lost on server restart
Closed
#45 [CRITICAL] Event system has 0% test coverage
Closed
#52 [MEDIUM] eval() in tools.py should use safer math evaluation
Closed
#47 [HIGH] XSS vulnerability in swarm_live.html (WebSocket data not sanitized)
Closed
#40 Fact distillation stores garbage and leaks secrets
Closed
#36 BUG: asyncio.run() in CLI kills event loop — interview and MCP tools fail
Closed
#39 Fix corrupted MEMORY.md and user_profile.md state
Closed
#27 Consolidate memory systems: 3 overlapping implementations
Closed
#11 Pending Actions Metadata Not Self-Correcting
Closed
#12 [Memory Discipline] Implement automatic session handoff at close to persist pending tasks and update status
Closed
#10 Working RAM Not Auto-Updating Timestamp
Closed
#23 Delete dead modules: openfang, paperclip, brain, swarm, cascade_adapter
Closed
#22 Delete dead code: agents/timmy.py (547 lines of superseded orchestrator)
Closed
#30 Remove or gate cloud integrations: Grok, Claude backends behind proper feature flags
Closed
#18 [SMOKE TEST] Final validation
Closed
#19 [SMOKE TEST] Module-level function
Closed
#16 [SMOKE TEST] StdioServerParameters integration
Closed
#17 [SMOKE TEST] create_gitea_issue_via_mcp
Closed
#15 [SMOKE TEST] MCP issue_write verification
41 Issues created by 3 users
Opened
#10 Working RAM Not Auto-Updating Timestamp
Opened
#11 Pending Actions Metadata Not Self-Correcting
Opened
#12 [Memory Discipline] Implement automatic session handoff at close to persist pending tasks and update status
Opened
#15 [SMOKE TEST] MCP issue_write verification
Opened
#16 [SMOKE TEST] StdioServerParameters integration
Opened
#17 [SMOKE TEST] create_gitea_issue_via_mcp
Opened
#18 [SMOKE TEST] Final validation
Opened
#19 [SMOKE TEST] Module-level function
Opened
#22 Delete dead code: agents/timmy.py (547 lines of superseded orchestrator)
Opened
#23 Delete dead modules: openfang, paperclip, brain, swarm, cascade_adapter
Opened
#24 Performance: eliminate redundant LLM calls in the request path
Opened
#25 Fix 62 bare except clauses across the codebase
Opened
#26 CLI: timmy chat breaks on multi-word messages and backticks
Opened
#27 Consolidate memory systems: 3 overlapping implementations
Opened
#28 Workspace correspondence: wire into heartbeat tick
Opened
#29 Replace 14 print() statements with proper logging
Opened
#30 Remove or gate cloud integrations: Grok, Claude backends behind proper feature flags
Opened
#31 agents.yaml: fix routing patterns and add debug route
Opened
#32 Docker cleanup: taskosaur references, root user, stale compose files
Opened
#36 BUG: asyncio.run() in CLI kills event loop — interview and MCP tools fail
Opened
#37 Memory consolidation: unify three stores into one coherent system
Opened
#38 Thinking engine: break the rumination loop
Opened
#39 Fix corrupted MEMORY.md and user_profile.md state
Opened
#40 Fact distillation stores garbage and leaks secrets
Opened
#45 [CRITICAL] Event system has 0% test coverage
Opened
#46 [HIGH] Chat history is in-memory only — lost on server restart
Opened
#47 [HIGH] XSS vulnerability in swarm_live.html (WebSocket data not sanitized)
Opened
#48 [MEDIUM] test_voice_loop.py import error — missing numpy dependency
Opened
#49 [MEDIUM] API endpoint naming is inconsistent — /api prefix usage varies
Opened
#50 [MEDIUM] Missing JSON status endpoints for key modules
Opened
#51 [LOW] Print statements should use proper logging
Opened
#52 [MEDIUM] eval() in tools.py should use safer math evaluation
Opened
#53 [LOW] Hardcoded model names in fallback chains should be configurable
Opened
#54 [HIGH] semantic_memory.py has zero test coverage
Opened
#55 [HIGH] Brain module severely under-tested (14.8% coverage)
Opened
#56 [MEDIUM] Test infrastructure has environment-specific failures
Opened
#64 [timmy-capability] Timmy has no awareness of his own session identity
Opened
#65 [timmy-capability] Timmy cannot run his own tests
Opened
#66 [timmy-capability] Timmy cannot read or create Gitea issues
Opened
#67 [timmy-capability] Timmy cannot delegate work to Kimi
Opened
#68 [timmy-capability] Timmy cannot reflect on his own past behavior