sync-ecosystem-data
Pull meeting transcripts, email threads, and calendar activity into /raw/ecosystem and index in /wiki. Use when syncing ecosystem data or as part of /data-ingestion.
SKILL.md.claude/skills/sync-ecosystem-data/SKILL.md
GitHubSync Ecosystem Data
Pipeline skill: pull personal-ecosystem signals into the knowledge base so the
system learns operational patterns. Runs standalone or as step 2 of
/data-ingestion.
Sources
Use whichever connectors are available this session (check with ToolSearch);
skip silently any that aren't connected:
| Signal | Tool |
|---|---|
| Meeting notes/transcripts | Notion notion-query-meeting-notes, Google Drive search |
| Email threads (non-newsletter) | Gmail search_threads — recent important threads |
| Schedule patterns | Google Calendar list_events — past 7 days |
| Document activity | Google Drive list_recent_files |
Process
- Find the last sync point. Read
raw/ecosystem/.last-sync; default to 7 days ago if missing. - Pull and digest. For each available source, fetch items since the sync point and extract only decisions, commitments, deadlines, recurring contacts, and operational patterns — not full bodies. Never store credentials, financial details, or obviously sensitive personal content; summarize around them.
- Write digests. One file per source with content:
raw/ecosystem/YYYY-MM-DD_<source>.md(e.g.2026-07-12_meetings.md). Skip sources with nothing new. - Index. Add one row per written file to the Ecosystem Data table in
wiki/index.md. - Update the sync point in
raw/ecosystem/.last-sync. - Report. Sources synced, sources skipped (and why), notable patterns.
If run from
/data-ingestion, keep this to 3 lines.