Skip to content
SkillModel-led3,669 tokens in use

wiki-agent

Searches one other agent's raw history for a topic, ingests only the matching sessions into your wiki, and returns a synthesized answer you can act on.

3,792 stars on the repository MIT LicenseAr9av/obsidian-wiki · skills/wiki-agent
pip install obsidian-wiki
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing wiki-agent costs

43tokens idle
3,669tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter8/15
  • Structure25/25
  • Examples15/15
  • Output15/15
  • Size13/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 254

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity100
  • Maintenance100
  • Community60
  • Maturity28
  • Momentum40

wiki-agent is a targeted, query-driven ingest. You are working in one agent and want context from another, so you call it with the target agent and a topic, for example a Codex or Pi question while you sit in Claude Code. It finds the sessions that match, extracts only the relevant slices, distills them into wiki pages and replies with a short synthesis.

Without a query it ingests the five most recent unprocessed sessions of that agent and summarizes them. Claude, Codex, Hermes, OpenClaw, Copilot and Pi histories are supported.

What You Get

  • A synthesized answer with key insights, the sessions it came from and the gaps it could not fill
  • New or updated wiki pages with sources recorded under an agent-prefixed locator
  • A manifest update for each processed session
  • A clear message when the history folder is missing or no session matches the query
  • A recent-sessions mode for catching up on one agent’s latest work

Common Use Cases

  • Asking Claude Code what you worked out about a topic in Codex last month
  • Pulling the context of a bug fixed in Hermes into the agent you use today
  • Catching up on the latest sessions of a single agent with no query

What Makes It Unique

It is built for crossing agents. Each agent gets its own extraction strategy, such as memory files first for Hermes and OpenClaw, rollout event types for Codex and active-branch walking for Pi. It reads sessions through grep windows so large files stay unread, shares its recency decay with the session graph skills so rankings agree, and finishes by answering the question, not only by filing pages.

Under The Hood

The skill is instructions for the agent. It reads history files from the other agent’s home folder on your machine, searches them with ripgrep and find, and writes wiki pages and the manifest into the vault. It runs the obsidian-wiki command line tool to log the run and refresh the index and hot cache, and the qmd command when a QMD collection is configured. The only folders it reads are the history roots you can override in your env file. It does not fetch web pages or call remote APIs.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers