Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From ai-memory
Manages AI memories: saves decisions/patterns/facts with categories/tags, searches context, lists/deletes entries, rescans project architecture via MCP tools and /remember /forget.
npx claudepluginhub damusix/ai-tools --plugin ai-memoryHow this skill is triggered — by the user, by Claude, or both
Slash command
/ai-memory:memory-managementThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
| Tool / Command | Purpose |
PROACTIVELY query Forgetful MCP (mcp__forgetful__* tools) when starting work on any project, when user references past decisions or patterns, when implementing features that may have been solved before, or when needing context about preferences. Save important decisions, patterns, and architectural insights to memory.
Manages persistent memory across Claude Code sessions via AutoMem. Recall project context, architectural decisions, bug fixes, user preferences, and patterns at session start or debugging.
Manages persistent semantic memory across sessions: store/retrieve knowledge/TODOs/issues, hybrid semantic search, hierarchy/tags organization, and maintenance tools.
Share bugs, ideas, or general feedback.
| Tool / Command | Purpose |
|---|---|
save_memory | Save a memory with category and tags |
search_memories | Full-text search across memories |
search_observations | Search atomic facts from turns |
list_memories | Browse memories with filters |
list_tags | List all tags with usage counts |
list_domains | List all domains with usage counts |
delete_memory | Delete a memory by ID |
rescan_project_architecture | Refresh tree + manifest snapshot for session context |
/remember | Slash command to save a memory |
/forget | Slash command to find and delete |
SAVE: Architectural decisions with rationale, non-obvious patterns, user preferences, discovered facts not in code/docs, solutions to tricky problems.
SKIP: Transient task details, things obvious from code, failed debugging steps, things findable by grep.
Pass as the category parameter to save_memory:
| Category | When |
|---|---|
| decision | Choice made between options, with rationale |
| pattern | Recurring approach established for codebase |
| preference | User style or workflow preference |
| fact | Discovered truth about system/environment |
| solution | Working fix for a non-obvious problem |
1=trivia, 2=useful, 3=normal (default), 4=important (confusion if forgotten), 5=critical (bugs/hours wasted if forgotten)
Use 2-5 tags per memory. Check existing tags with list_tags first — prefer existing vocabulary. Good tags: domain (auth, database), technology (react, postgres), pattern type (architecture, testing), component names (user-service, checkout).
FTS5 syntax: "exact phrase", term* for prefix, term1 OR term2. Try multiple queries if first returns nothing. Check list_tags to discover what exists.
For detailed FTS5 query syntax and search strategies, see references/search-patterns.md.