← ClaudeAtlas

mem-searchlisted

Search claude-mem's persistent cross-session memory database. Search for an index, filter, then fetch full details only for kept IDs, for 10x token savings. Use when the user asks "did we already solve this?", "how did we do X last time?", or needs work from previous sessions. Requires the claude-mem MCP.
HermeticOrmus/mem-search-skills · ★ 0 · AI & Automation · score 75
Install: claude install-skill HermeticOrmus/mem-search-skills
# Mem Search Recall past work across all sessions cheaply. Workflow: search, filter, fetch. **Requires the claude-mem MCP.** This skill assumes the `search`, `timeline`, and `get_observations` tools are available. If they are not, claude-mem is not installed. ## When to use Use when the user asks about PREVIOUS sessions, not the current conversation: - "Did we already fix this?" - "How did we solve X last time?" - "What happened last week?" If the answer is in the current conversation, just answer. ## The 3-layer workflow (always follow) Never fetch full details without filtering first. The index is cheap, full observations are expensive: roughly 10x token savings by filtering before fetching. ### Step 1: search, get the index with IDs Use the `search` tool: ```text search(query="authentication", limit=20, project="my-project") ``` Returns a table of IDs, timestamps, types, titles (~50-100 tokens per result): ```text | ID | Time | Type | Title | Read | |--------|---------|--------|-----------------------------|------| | #11131 | 3:48 PM | obs | Added JWT authentication | ~75 | | #10942 | 2:15 PM | bugfix | Fixed auth token expiration | ~50 | ``` Parameters: - `query` (string) search term - `limit` (number) max results, default 20, max 100 - `project` (string) project name filter - `type` (string, optional) "observations", "sessions", or "prompts" - `obs_type` (string, optional) comma-separated: bugfix, feature, decision, d