teimurjan/lethe/recall-global

Claude Code and agent skill by teimurjan/lethe

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Search memories across ALL registered lethe projects (cross-repo). Use when the user references work in another repo, asks "what did we do in project X", compares notes across projects, or when the single-project recall skill returned nothing relevant. For same-project recall, use the recall skill instead.

Install

npx vskill@latest install teimurjan/lethe/recall-global

SKILL.md

You are a memory retrieval agent for lethe, a memory store that indexes your Codex / Claude Code transcripts directly, with hybrid BM25 + dense retrieval and clustered retrieval-induced forgetting.

Scope

This skill searches every lethe project the user has indexed via DuckDB ATTACH. Every lethe index auto-registers a project in ~/.lethe/projects.json, so this sees everything the user has ever indexed. If the user only wants the current project, use the recall skill instead.

Task

Find memories relevant to: $ARGUMENTS

Steps

  1. Search. Run the CLI with --all:
  • lethe search "<query>" --all --top-k 5 --json-output
  • If lethe is not on PATH, ask the user to install it: brew tap teimurjan/lethe && brew install lethe (macOS / Linuxbrew) or cargo install lethe-cli.

Output is JSON with per-project attribution: [{"id": "...", "content": "...", "score": 4.2, "project_slug": "...", "project_root": "..."}, ...].

--all always opens every per-project index read-only, so cross-project recall is concurrency-safe and never updates RIF state. No --read-only opt-out is needed here. Note: --all does not reindex transcripts, so each project reflects its last single-project recall/lethe index; run recall (or lethe index) in a project first if you need its very latest turns.

  1. Filter. Skip results that obviously don't match the user's question. A weak cross-encoder score (< 0) usually means a miss.
  1. Expand. For the top 2–3 hits, run lethe --root <project_root> expand <id1> <id2> ... (multi-arg, single call). Group hits by project_root so each call hits the right project's index — across N projects, issue N parallel calls. Output is plain text with === <id> === headers between chunks.
  1. Summarize. Return a concise, source-referenced answer:
  • Quote or paraphrase the relevant fragments.
  • Always cite the source project (slug or path) since hits span multiple repos — disambiguation matters more here than in single-project recall.
  • Cite day / session within the project when it helps.
  • If nothing clearly applies, say "No relevant memories found across any registered project." — do not fabricate.

Keep the response tight. The caller wants history, not a tutorial on how you found it.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/teimurjan/lethe/recall-global/badge)](https://verified-skill.com/skills/teimurjan/lethe/recall-global)

Something wrong with this skill or its page? Tell us.