copilot-history-ingest turns your GitHub Copilot history into wiki knowledge. It reads three places: the per-session folders under ~/.copilot/session-state, the global session-store database, and the Copilot chat data in VS Code workspace storage.
Checkpoints and session summaries are ingested first because Copilot already distilled them. Raw turns come next, then saved plans and file and commit references, all grouped by topic and written with provenance markers.
What You Get
- Project pages named after each repository, with architecture notes and important files
- Concept, skill and entity pages distilled from checkpoints and turns
- Links between wiki knowledge and the commits, PRs and issues recorded in sessions
- A manifest of ingested sessions, checkpoints and memory artifacts
Common Use Cases
- Capture what Copilot CLI sessions taught you into a vault
- Import plans saved in VS Code Copilot chat as wiki pages
- Find the files you touch most per repository from session records
- Sync only sessions updated since the last ingest
What Makes It Unique
It covers three Copilot data sources at once and gives SQL queries for the session database, including checkpoints, file touch counts and session references. It tells the agent to skip internal reasoning fields entirely and to treat user-saved plan files as high quality input.
Under The Hood
The skill is instructions for the agent. It tells the agent to read the vault config and manifest, list and read the Copilot folders, run SQL queries against the local session-store SQLite database, decode a base64 folder name with a short Python snippet, and call the obsidian-wiki command line to sync the index, log and hot cache. It writes markdown pages and manifest entries into your vault, and refreshes a QMD search index when one is configured. It makes no web requests of its own.