Productivité

agent-recall

Provides persistent, local memory for AI agents through MCP tools for session context, recall, corrections, decision tracking, cross-project insights, and journal consolidation.

★ 371 GitHub Stars
94 Score de qualité
1 Fichiers

About this Skill

Persistent compounding memory for AI agents. 5 default MCP tools: session_start, session_end, remember, recall, check. Full surface (18 tools) available with --full flag. Two-verb model: inhale (session_start) and exhale (session_end). Correction-first memory with decision trail tracking, watch_for warnings, palace rooms with salience scoring, cross-project insight matching, same-day journal merging, ambient recall hooks. Local markdown only. Zero cloud, zero telemetry, Obsidian-compatible. Optional Supabase backend: when configured via `ar setup supabase`, recall() uses pgvector cosine similarity on OpenAI/Voyage embeddings instead of keyword search — same API, semantic understanding. Gracefully degrades to local search if not configured.

Cas d’usage

  • Load context at session start
  • Save session summaries and insights
  • Recall decisions and lessons
  • Track corrections and decision quality
  • Share relevant insights across projects

Fonctionnalités

  • Five default MCP tools with an optional 18-tool full mode
  • Local Markdown storage compatible with Obsidian
  • Correction tracking with predictive watch_for warnings
  • Cross-project insight matching and salience-based palace rooms
  • Optional Supabase pgvector semantic search

Prérequis

  • Node.js
  • npx
  • An MCP-compatible agent

Agents compatibles

Claude Code Codex Cursor GitHub Copilot Windsurf Roo Code Other

Skill files

Related Skills

markit

Converts local files, URLs, GitHub resources, and other supported content into Markdown through a CLI or TypeScript SDK.

shift-labs-ai/markit ★ 1324
markdown document conversion file extraction pdf

total-recall

Provides autonomous cross-session memory for OpenClaw by observing conversation transcripts, compressing them into prioritized notes, consolidating large logs, and recovering miss…

gavdalf/total-recall ★ 273
agent memory conversation memory openclaw session recovery

resume-jd-optimizer-cn

Analyzes Chinese job descriptions and resumes, maps evidence and gaps, asks targeted questions, rewrites tailored resumes, and prepares ATS, HR, interview, and recruiting communic…

coinluu/resume-jd-optimizer-cn ★ 184
resume job description career ats

lets-go-rss

Aggregates updates from multiple social and media platforms, supports incremental deduplication, AI classification, scheduled refreshes, cached status reports, feed exports, and h…

ALBEDO-TABAI/lets-go-rss ★ 102
rss content aggregation subscriptions social media

jira

Provides a terminal CLI for Jira Cloud issue searches, details, creation, comments, project listing, approximate counts, transitions, and status changes, with JQL and REST API gui…

magnus919/agent-skills ★ 88
jira atlassian issue-tracking jql

notion

Enables agents to search, retrieve, query, and update Notion pages and databases through a bounded CLI, with dry-run and explicit confirmation safeguards for mutations.

magnus919/agent-skills ★ 88
notion knowledge base pages databases

anydoc

Converts office documents, spreadsheets, presentations, ebooks, CSV files, and PDFs to GitHub-Flavored Markdown locally using the pinned Any Doc CLI.

magnus919/agent-skills ★ 88
document conversion markdown office files pdf

life-coach

Guides bounded, user-led AI coaching for personal goals, decisions, habits, transitions, accountability, and progress review while enforcing nonclinical scope and safety referrals.

magnus919/agent-skills ★ 88
life coaching goal setting habits accountability