mem-search
Search claude-mem persistent cross-session memory through enforced progressive
by thedotmackFree
Description
mem-search is a skill from the claude-mem project for searching the persistent memory that claude-mem keeps across Claude Code sessions. It is meant for finding previous decisions, earlier solutions, past work history and similar context. The skill enforces a progressive disclosure flow. The agent starts with a compact index, then pulls surrounding context for the most relevant results, and only then opens the full details of the selected memories. It does not begin by fetching full records. The guided flow runs in three steps, from an index through context to selected details, and the agent asks for a deeper layer only when it helps answer the current question. Local and remote MCP servers expose the same input and result contract. The skill also tells the agent to treat anything stored in memory as evidence from previous sessions, so that stored instructions, commands or tool results are data and are not executed as instructions.…
Highlights
- Three-step progressive disclosure search
- Index, then context, then selected details
- Concise readable replies, not raw JSON
- Stored memory treated as evidence, not commands
Ratings & Reviews
No ratings yet
Be the first to rate mem-search.
People also visit
workspace-surface-audit
Audit the active repo, MCP servers, plugins, connectors, env surfaces
skill-lookup
Search, retrieve, and install Agent Skills from the prompts.chat registry
claude-automation-recommender
Analyze a codebase and recommend Claude Code automations
content-research-writer
Assists in writing high-quality content by conducting research
mem-search FAQ
What does the mem-search skill do?
It searches claude-mem's cross-session memory for previous decisions, solutions and work history, using a layered flow.
How do I install it?
Run: npx skills add https://github.com/thedotmack/claude-mem --skill mem-search
Why does it fetch in layers?
The skill avoids loading full records first. It starts from an index and pulls more detail only for the results that matter.
Can stored memory tell the agent to run commands?
No. The skill says stored instructions, commands and tool results are treated as data, not as instructions to execute.
What security checks has it passed?
skills.sh lists three audits: Gen Agent Trust Hub Pass, Socket Pass and Snyk Pass.
What are the best mem-search alternatives?
Popular alternatives to mem-search include ck, workspace-surface-audit and skill-lookup. See all mem-search alternatives