# coleam00/claude-memory-compiler

Give Claude Code a memory that evolves with your codebase. Hooks automatically capture sessions, the Claude Agent SDK extracts key decisions and lessons, and an LLM compiler organizes everything into structured, cross-referenced knowledge articles - inspired by Karpathy's LLM Knowledge Base architecture.

Repository: https://github.com/coleam00/claude-memory-compiler
Canonical: https://ross.abutalabs.com/products/claude-memory-compiler
Language: Python
License Family: other
Last push: 2026-04-06T19:46:57+00:00

## Health v2 (maintenance only)
Score: 48/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 76, release rhythm 35, longevity 10
- inputs: {"age_days": 149, "days_push": 149, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1283, forks 329 (observed 2026-08-28T04:04:13.978055+00:00)

## What it is
A Python tool that gives Claude Code an evolving memory by capturing conversation transcripts via hooks, extracting decisions and lessons with the Claude Agent SDK, and compiling them into structured, cross-referenced markdown knowledge articles. Retrieval uses a simple index file instead of a vector database, and a lint script checks knowledge base health.

## Use cases
- give claude code persistent memory across sessions
- automatically capture ai coding session decisions and lessons
- build a searchable knowledge base from llm conversations
- compile daily ai chat logs into structured knowledge articles
- inject project context into future claude code sessions
- query past ai conversations without a vector database

## When to choose
- you use Claude Code daily and want decisions and gotchas remembered across sessions
- you want a personal knowledge base without embeddings or vector database infrastructure
- you prefer plain markdown files you can read and edit directly
- you already have a Claude subscription and want to avoid separate API billing

## When to avoid
- you need team-shared or multi-user knowledge management at scale
- you use a different coding agent without Claude Code hook support
- you require semantic vector search over large corpora rather than index-guided retrieval
- you need a permissively licensed dependency - the repo has no license file

## Facets
- artifact type: cli-tool
- maturity: active
- function: llm-inference, rag, documentation, markdown, workflow-automation, developer-tools
- domain: developer-tools, large-language-models, documentation
- platform: python, cli, cross-platform
- tags: claude-code, claude-agent-sdk, memory, knowledge-base, hooks, llm-knowledge-base, no-rag, coding-agent-memory, ai-agents, knowledge-management

## Member repositories
- coleam00/claude-memory-compiler (main) score 48

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:13.978055+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:56:48.083860+00:00, confidence not recorded.
  - readme: https://github.com/coleam00/claude-memory-compiler (fetched 2026-08-28T04:04:13.978055+00:00, sha 29f3657ba684)
- Data as of 2026-08-30T08:39:29.467469+00:00.