Skip to content

Recall Past Memories

Searches an agent's saved memory of past sessions for a topic and reports only what it finds.

Stars
30,083
Licence
Apache-2.0
Files
3 files

What it does

Once installed, when you ask the agent to recall something (for example "what did we do about refresh tokens?"), it searches a persistent memory store called agentmemory for matching past observations, decisions and session notes. It groups the results by session, leads with the most important ones, and is instructed to quote only what the tool returned, suggesting alternative search terms instead of guessing when nothing matches.

Great for

  • Developers who use agentmemory to keep context across coding sessions and want to look up past decisions
  • Teams reviewing why a previous technical choice was made
  • Users who want the agent to avoid inventing history when memory is empty

Skip it if

  • People who have not installed the agentmemory tool: the skill only calls its search function and will report it unavailable otherwise
  • Situations needing guaranteed-accurate records from outside sources, since results are only as good as what was previously saved

Try asking

  • “recall what we decided about jwt refresh token rotation”
  • “did we ever fix the rate limiter per-IP bug?”
  • “have we seen this database timeout before?”
  • “recall the session where we discussed the billing migration”
Step-by-step guide: how to use Recall Past MemoriesInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the Apache-2.0 licence and an attribution note.

    recall.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip recall.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip recall.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use specific keywords or phrases, since the search combines keyword, vector and graph matching and works best with precise terms
  • Scope to a repo or agent when you have several projects so results are not mixed together
  • If results are empty, try the alternative terms the skill suggests before concluding nothing was recorded

What's inside

  • EXAMPLES.md
  • SKILL.md
  • TROUBLESHOOTING.md

Redistributed unmodified from rohitg00/agentmemory under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.