recall

Search per-project memory stores using BM25 ranking and configurable filters.

Updated Aug 27, 2026

Search Tags:#context #recall #ai-assistant #project-memory #worklog #memory-search #bm25

Ready to run

npx skills add https://github.com/calebeynon/claude-code-setup --skill recall-calebeynon

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: recall
Source: https://github.com/calebeynon/claude-code-setup/tree/main/skills/recall
Command: npx skills add https://github.com/calebeynon/claude-code-setup --skill recall-calebeynon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Quickly surface relevant past context from per-project memories and archived sessions to inform current work and decisions.

Core Features & Use Cases

Memory search: Query memories stored for the current project to retrieve relevant chats, worklogs, and decisions.
Context-aware recall: Filter results by type (project, feedback, decision, gotcha, worklog, reference) and constrain scope to the current project or across all projects.
Rapid briefings: Present brief previews of top results to guide next steps without exposing sensitive data.

Quick Start

Use recall to fetch relevant memories and display top results to inform your next action.

Frequently Asked Questions about recall

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I search past project memory and worklogs to find context for current work?▼

Project memory search uses BM25 ranking to query per-project memory stores, surfacing relevant past chats, worklogs, and decisions to inform current tasks. It filters results by type and project scope.

What is the best way to retrieve archived AI assistant chats for a specific project?▼

Retrieving archived AI assistant chats is done by querying the per-project memory store under the current project key, returning top ranked results with brief previews to guide next steps.

Can I filter project memory search results by type, like decisions or feedback?▼

Project memory search supports configurable filters by type, allowing you to narrow results to decisions, feedback, gotchas, worklogs, references, or general project context across stored memories.

How does BM25 ranking work for surfacing relevant project context?▼

BM25 ranking scores stored memory entries against your search query, returning the top results based on term frequency and relevance. It surfaces the most relevant past context for current work.

Do I need a specific script to run project memory recall?▼

Project memory recall requires the memory_search.py script to run. It operates against memories stored under the project key directory to search and surface stored context.

Can I search across all projects or is memory recall scoped to the current project only?▼

Memory recall is scoped to the current project by default, but includes configurable filters to expand the search scope across all projects if needed for broader context retrieval.