Clarifies when to use native Chrome integration vs playwright-cli
(headless automation, network mocking, persistent sessions, SSH/containers).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Extract API and CLI references into dedicated files:
- SKILL.md (173 lines): routing, safety rules, architecture, common queries
- API_REFERENCE.md (99 lines): endpoints, auth, Python client
- CLI_REFERENCE.md (115 lines): commands, flag ordering, compliance workflow
Total content reduced from 1005 to 387 lines.
Context loaded per session drops from 1005 to 173 lines.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- create-skill: superseded by native /skill command and official docs
- orchestrator: superseded by native Agent Teams (/agent-teams)
- notediscovery: redundant with cognitive-memory MCP (superior local-first implementation)
All moved to _archive/ for historical reference.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Rewrite issue-worker agent with phased workflow, self-review step,
label management, memory storage, and test discovery
- Update .mcp.json configuration
- Remove stale mcp-needs-auth-cache.json
- Update plugins submodule and blocklist/marketplaces
- Minor update to mcp-manager skill
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- backlog/SKILL.md: Prioritize gitea-mcp MCP server over raw curl API calls
- settings.json: Updated configuration
- plugins: Updated blocklist and known_marketplaces
- Removed stale mcp-needs-auth-cache.json
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Moved SCOPE restrictions from frontmatter description to markdown body to reduce
character budget consumption. Added description budget best practices to create-skill.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
New skill for creating/managing headless Claude scheduled tasks
on systemd timers. Plugin blocklist and marketplace files updated.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Multi-graph: named graph routing in MCP server (graph param on all tools),
CLI --graph flag, graphs subcommand, resolve_graph_path() in common.py
- XDG compliance: data dir resolves via COGNITIVE_MEMORY_DIR env > XDG_DATA_HOME > ~/.local/share/
- Remove CORE.md auto-loading: drop MEMORY.md symlinks, update CLAUDE.md to MCP-first recall
- Update all scripts (git-sync, ensure-symlinks, edge-proposer) for portable path resolution
- Remove symlinks step from daily systemd service
- Version bump to 3.1.0
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
3,348-line monolith → 6 modules with mixin classes resolving via MRO.
client.py retains __init__, internal helpers, and core CRUD (1,091 lines).
All backward-compat imports preserved for mcp_server.py and dev/migrate.py.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Closescal/claude-memory#3. All ID-based lookups now resolve partial
UUID prefixes (git-style) instead of requiring exact full UUIDs.
Affects _resolve_memory_path, _resolve_edge_path, edge_search, and
related(). Ambiguous prefixes raise ValueError with candidates.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Daily and weekly timers fired simultaneously on Sundays, causing
decay() and reflect() to race on _state.json. Now merges top-level
keys before writing and uses atomic tempfile+rename to prevent
partial reads from triggering silent JSONDecodeError fallbacks.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- .gitignore: Add backups/ to stop noisy session backup rotation diffs
- Remove tracked backup files from index
- skills/cognitive-memory/mcp_server.py: Auto-push to Gitea on memory_store and memory_relate
- skills/cognitive-memory/scripts/memory-git-sync.sh: New git commit+push script
- skills/cognitive-memory/systemd/cognitive-memory-daily.service: Add git sync as fallback
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- SKILL.md: New workflow pattern for creating edges between memories
- SKILL.md: Added proactive trigger for edge creation after storing memories
- SKILL.md: Updated bug fix example to show relating solution to problem
- scripts/edge-proposer.py: Retroactive edge candidate analysis tool
Scores memory pairs by type heuristics, tag overlap, and content similarity
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- commands/save-memories.md: slash command that analyzes session context,
finds cutoff point from last memory save, gates on value, and spawns
memory-saver agent in background with structured summary
- agents/memory-saver.md: sonnet-based agent that stores cognitive memories
from structured summaries via MCP tools or CLI fallback
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
MCP store was missing the --episode behavior that the CLI had,
so episodes stopped being logged when sessions switched to using
MCP tools. Now defaults to true with opt-out via episode=false.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Normalize both signals to 0-1 range so semantic similarity scores
aren't drowned out by keyword position scores. Jellyfin DB recovery
now ranks #1 for "media server database broken sqlite error" instead
of being buried behind keyword noise matches on "error".
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- embed --if-changed compares memory ID sets to skip when unchanged
- --ollama-model config flag to switch models without code changes
- _ollama_embed now accepts model param, respects config throughout
- Switched default to qwen3-embedding:8b (70+ MTEB vs nomic's ~52)
- Hourly systemd timer for embedding, daily timer no longer embeds
- Fixed PATH for uv in both systemd services
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
