obsidian-wiki

Install

pip install obsidian-wiki
obsidian-wiki setup --vault ~/brain

This is third-party code. Review the repository files before installing.

What it does

The project follows Andrej Karpathy's LLM Wiki pattern: instead of re-answering the same questions, the agent compiles what it learns into linked markdown pages in an Obsidian vault. Skills ingest folders, URLs, PDFs, the current repository and conversation history from Claude Code, Codex, Copilot and other agents. Queries return answers with wikilink citations, and maintenance skills find broken links, duplicates and contradictions and weave new pages into the graph. A manifest tracks what was ingested, so later runs process only the delta.

Who it is for. For developers and researchers who want to keep knowledge from agent sessions in their own base instead of losing it in chat logs.

Example request

Save this session to the wiki and link it to what I already know about rate limiting

Limitations

Ingesting large chat archives costs many tokens; the author suggests pre-extracting JSONL with the bundled script. Page quality depends on the agent's model. Python is required for the CLI and helper scripts.

How to disable. Remove the package with pip uninstall obsidian-wiki and delete the .skills symlinks from your agents' skill directories. The vault stays a plain markdown folder.

Security check

• Reads agent conversation history and project files
• Writes pages to the local vault and runs Python scripts

README in short

The README explains the idea: knowledge is compiled once and kept current rather than retrieved again through RAG. Install via pip, uv or pipx and run setup with a vault path, or ask the agent to do it. It lists commands for ingesting, querying, digests and wiki cleanup, plus a CLI for searching past sessions. The graph exports to GraphML, Neo4j, SQL or interactive HTML. MIT licensed.

SKILL.md

---
name: claude-history-ingest
description: >
  Ingest Claude Code conversation history into the Obsidian wiki. Use this skill when the user wants to mine
  their past Claude conversations for knowledge, import their ~/.claude folder, extract insights from
  previous coding sessions, or says things like "process my Claude history", "add my conversations to the wiki",
  "what have I discussed with Claude before". Also triggers when the user mentions their .claude folder,
  Claude projects, session data, past conversation logs, local-agent-mode sessions, or audit logs.
---

# Claude History Ingest — Conversation Mining

You are extracting knowledge from the user's past Claude Code conversations and distilling it into the Obsidian wiki. Conversations are rich but messy — your job is to find the signal and compile it.

This skill can be invoked directly or via the `wiki-history-ingest` router (`/wiki-history-ingest claude`).

## Before You Start

1. **Resolve config** — follow the Config Resolution Protocol in `llm-wiki/SKILL.md` (inline `@name` override → walk up CWD for `.env` → global config → prompt setup). This gives `OBSIDIAN_VAULT_PATH` and `CLAUDE_HISTORY_PATH` (defaults to `~/.claude`)
2. Read `.manifest.json` at the vault root to check what's already been ingested
3. Read `index.md` at the vault root to know what the wiki already contains

## Ingest Modes

### Append Mode (default)

Check `.manifest.json` for each source file (conversation JSONL, memory file). Only process:

- Files not in the manifest (new conversations, new memory files, new projects)
- Files whose modification time is newer than their `ingested_at` in the manifest

Related