skills.rest
oresttokovenko
oresttokovenkoCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

claude-history-ingest

Extracts knowledge from Claude Code conversation history into an Obsidian wiki.

Updated Nov 9, 2023
Search Tags:#claude code#conversation history#obsidian#knowledge extraction#personal knowledge management#jsonl parsing#wiki ingest
One-click install
npx skills add https://github.com/oresttokovenko/dot-files --skill claude-history-ingest-oresttokovenko
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: claude-history-ingest
Source: https://github.com/oresttokovenko/dot-files/tree/main/chezmoi/dot_agents/skills/claude-history-ingest
Command: npx skills add https://github.com/oresttokovenko/dot-files --skill claude-history-ingest-oresttokovenko

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Past Claude Code sessions contain valuable decisions, debugging patterns, and project context, but they sit unread in noisy JSONL logs. This Skill mines those conversations and distills them into structured, searchable Obsidian wiki pages. ## Core Features & Use Cases - Incremental ingest: Tracks processed files in a manifest so only new or modified conversations, memory files, and audit logs are re-processed. - Multi-source parsing: Reads CLI sessions under ~/.claude, desktop local-agent-mode sessions, memory files, and audit logs, preferring pre-extracted compact JSON over raw JSONL. - Topic clustering and distillation: Groups knowledge by topic across conversations and writes project, concept, skill, and synthesis pages with provenance and confidence metadata. - Use Case: After weeks of Claude-assisted work, run the ingest to turn scattered sessions into a personal knowledge base with project overviews, recurring debugging patterns, and workflow notes. ## Quick Start Process my Claude history and add the distilled knowledge from my past conversations to the Obsidian wiki.

Frequently Asked Questions about claude-history-ingest

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

FAQPage Schema
How do I import my Claude Code conversation history into Obsidian?▼

Run the ingest against your ~/.claude folder. It scans projects, memory files, and session JSONL transcripts, then distills the content into topic-clustered wiki pages in your Obsidian vault while recording progress in a manifest.

How does incremental ingest avoid reprocessing old conversations?▼

A .manifest.json file at the vault root records each source file with its ingestion timestamp and modification time. On append mode, only files that are new or modified since their recorded ingest are processed.

Can I exclude certain projects from the history ingest?▼

Yes. Set WIKI_SKIP_PROJECTS in the resolved config as a comma-separated list of substrings. Any project directory whose name contains one of those substrings is excluded from scanning, sampling, and manifest writes.

Does it support Claude desktop app sessions, not just the CLI?▼

Yes. It scans ~/Library/Application Support/Claude/local-agent-mode-sessions for session metadata, audit.jsonl logs, and conversation transcripts, processing them alongside CLI data from ~/.claude.

Why is pre-extraction recommended before ingesting raw JSONL?▼

Raw JSONL files are mostly tool_use, thinking, and progress noise. The extract-jsonl.py helper strips that down to user and assistant text turns, achieving 50-200x size reduction so far more conversations fit in the token budget.

What happens to sensitive data like API keys in my conversations?▼

The skill distills knowledge rather than copying raw conversation text, and explicitly skips anything resembling secrets, tokens, passwords, or credentials. Personal or sensitive content is flagged for user confirmation before inclusion.

Related Skills

Explore More Personal & Entrepreneur Skills→
itsmarsssitsmarsss
4

claude-history-ingest

Ingest Claude Code conversation history into structured Obsidian wiki pages.

Community
Advanced
NaoyaMiyagawaNaoyaMiyagawa
1

custom-obsidian-wiki-query

Query an Obsidian knowledge vault with qmd hybrid search to answer domain and codebase questions.

Community
Intermediate
NirajkashyapNirajkashyap

wiki-ingest

Ingest documents into an Obsidian wiki as interconnected wiki pages.

Community
Advanced
Michael1PengMichael1Peng

openclaw-history-ingest

Ingest OpenClaw agent history into Obsidian wiki pages with provenance tagging.

Community
Advanced