raiyanyahya/recall

Stop wasting tokens and re-explaining your project every session. Recall gives Claude Code , Opencode durable memory — entirely offline.

What it solves

Recall provides a fully-local, automatic session memory for AI coding agents (specifically Claude Code and OpenCode), solving the "cold-start" problem where users must re-explain project context, goals, and progress to the AI in every new session. It reduces token usage and costs by replacing long transcripts with a compact, locally-generated summary.

How it works

Recall captures session activity (prompts, replies, files touched, and commands run) into an append-only history.md file. When a user saves the context, a local Python summarizer uses TF-IDF and TextRank algorithms to extract the most central sentences from the history to create a context.md file. This summary includes the project goal, files touched, and git diffs, which is then loaded into the next session to resume work.

Who it’s for

Developers using Claude Code or OpenCode who want to maintain persistent project memory without sending their session transcripts to an external API or incurring additional LLM costs for summarization.

Highlights

  • Fully Local: No API keys or external models; summarization is done via a classical Python algorithm.
  • Token Efficient: Resumes sessions from a compact digest (~1–2K tokens) rather than replaying full transcripts.
  • Privacy-First: Session transcripts and secrets are kept on the local machine with built-in redaction for sensitive data.
  • Harness Agnostic: While built for Claude Code, it uses plain markdown files, allowing it to be shared across different coding agents.
  • Zero Dependencies: The summarizer is vendored and runs on Python stdlib, with NumPy as an optional accelerator.

Related

  • Dispatch
    Recall: Local Project Memory for Claude Code

    Recall is an offline, local-first plugin for Claude Code that captures session history and uses a TF-IDF and TextRank summarizer to create durable project memory without spending model tokens.

  • Project
    supermemoryai/claude-supermemory

    claude-supermemory is a Claude Code plugin that gives the LLM coding agent persistent memory across sessions. It automatically captures conversation turns, recalls relevant past context at prompt time, and shares that memory across tools like Codex and OpenCode via a unified container tag.

  • Project
    EliaAlberti/cpr-compress-preserve-resume

    A Claude Code plugin that provides persistent memory across sessions by compressing conversations into curated logs and automatically restoring context.

  • Project
    Ancienttwo/repo-harness

    A file-backed workflow and runtime for AI agents that replaces chat memory with repository-based state, enabling resumable sessions and token-efficient development.

  • Project
    akitaonrails/ai-memory

    ai-memory is a long-term memory system for AI coding agents, storing memory as git-backed Markdown, working across agents, projects, machines, and teams, with zero-LLM default and optional LLM features.