agentleFS
thieungthieung/agent-skills/skills/recall/SKILL.md

recall

Skill4 starsChanged 9 months agoMITView on GitHub

Load context from Claude Code session history. Temporal queries use native JSONL timelines, topic queries use QMD BM25 search, and "recall graph" creates an interactive session-file graph. Use when users say "recall", "what did we work on", "load context about", "remember when we", "prime context", "yesterday", "what was I doing", "last week", "session history", "recall graph", "session graph", or ask to visualize session relationships.

  • Installs packages

142 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use addskill, apply-firmware and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • ryanportfolioSkill

    addskill

    Create, import, update, or install repo or personal skills: runtime ownership, resources, discovery checks.

    Also in 6 projects that use this

  • ryanportfolioSkill

    apply-firmware

    Bring current folder (plain folder, existing repo, old or partial firmware copy) up to full Harness Firmware in place. Mirror external repo → adopt-repo.

    Also in 6 projects that use this

  • +10 more from ryanportfolio

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 129 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every recall response should end with One Thing when there is enough signal: the single highest-leverage next action synthesized from the …

    line 19
  2. Distinctive:

    Do not assume cwd is the skill directory.

    line 23
  3. Distinctive:

    Use QMD BM25 search. Expand the user query into 3-4 keyword variants because BM25 needs exact-ish terms.

    line 71
  4. Very specific:

    Read workflows/recall.md for routing logic and response shape.

    line 106

“Common” is measured: kept by at least 6 in 10 of the 129 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Recall

Recall has three modes:

  • Temporal: date-based session timeline from Claude Code JSONL files.
  • Topic: BM25 search across QMD collections.
  • Graph: interactive HTML graph of sessions and files touched.

Every recall response should end with One Thing when there is enough signal: the single highest-leverage next action synthesized from the results.

Script Paths

Do not assume cwd is the skill directory. Resolve the installed skill path first:

RECALL_DIR="/path/to/recall"
python3 "$RECALL_DIR/scripts/recall-day.py" list yesterday

Common installed paths:

  • Global Claude Code skill: ~/.claude/skills/recall
  • Project skill: .claude/skills/recall
  • This repository: skills/recall

Configuration

Defaults work for standard Claude Code installs:

VariableDefaultPurpose
CLAUDE_HOME~/.claudeClaude Code home
CLAUDE_PROJECTS_DIR$CLAUDE_HOME/projectsJSONL session directory
VAULT_DIRdetected from cwd if .obsidian existsObsidian vault root, used by graph mode to normalize file paths

Topic search uses the QMD collection claude-sessions maintained by the session-sync skill. Install and index session-sync first (see "QMD Index" below).

Graph mode requires Python packages in requirements.txt:

python3 -m pip install -r "$RECALL_DIR/requirements.txt"

Temporal Recall

python3 "$RECALL_DIR/scripts/recall-day.py" list yesterday
python3 "$RECALL_DIR/scripts/recall-day.py" list "last week"
python3 "$RECALL_DIR/scripts/recall-day.py" list 2026-02-25
python3 "$RECALL_DIR/scripts/recall-day.py" expand SESSION_ID

Options:

  • --min-msgs N: filter short/noisy sessions.
  • --all-projects: scan all projects.
  • --project PATH: scan one project directory or source path.

Topic Recall

Use QMD BM25 search. Expand the user query into 3-4 keyword variants because BM25 needs exact-ish terms.

qmd search "VARIANT_1" -c sessions -n 5
qmd search "VARIANT_2" -c sessions -n 5
qmd search "VARIANT_1" -c notes -n 5
qmd search "VARIANT_1" -c daily -n 3
qmd get "qmd://collection/path/to/file.md" -l 50

Deduplicate by document path and present the top unique results.

Graph Recall

python3 "$RECALL_DIR/scripts/session-graph.py" yesterday --no-open
python3 "$RECALL_DIR/scripts/session-graph.py" "last week" --min-files 5 --no-open -o "$RECALL_DIR/output/session-graph.html"

Tell the user the output path, node/edge counts, and what clusters/shared files indicate.

QMD Index

The QMD collection claude-sessions is owned by session-sync. Run once:

SS="$HOME/.claude/skills/session-sync/scripts/session-sync.py"
python3 "$SS" export --all
python3 "$SS" index

The Stop hook (see session-sync/procedures/setup-hook.md) keeps the collection fresh automatically.

Workflow

Read workflows/recall.md for routing logic and response shape.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.