Persisting Bigpicture Learnings

Persist Bigpicture to Learnings Hub

Target: $ARGUMENTS

Persists the output of /synthesizing-cc-bigpicture as a dated snapshot in a learnings hub. Maintains a latest.md pointer (always current) and an append-only archive for compound cross-session learning.

Arguments

Examples:

/persisting-bigpicture-learnings
/persisting-bigpicture-learnings ~/.claude/bigpicture.md
/persisting-bigpicture-learnings bigpicture.md docs/learnings/cc-bigpicture/

Output Structure

docs/learnings/cc-bigpicture/
├── latest.md          # Always the most recent synthesis
└── archive/
    ├── 2026-04-01.md
    ├── 2026-04-03.md
    └── 2026-04-06.md  # Today's snapshot

Workflow

1. Parse arguments — Apply defaults per Arguments table.

2. Read input — Read input-path. If the file does not exist, report error and stop.

3. Create hub directory — Create hub-path and hub-path/archive/ if absent.

4. Write latest — Write full synthesis content to hub-path/latest.md. Always overwrite — this is the current-state pointer.

5. Write archive snapshot — Determine today's date (YYYY-MM-DD).

   • If hub-path/archive/YYYY-MM-DD.md does not exist, create it with the synthesis content.
   • If it does exist, append a --- separator followed by the new synthesis content. This preserves all snapshots from the same day.

6. Report — Confirm paths written and archive entry count for today.

Integration

Runs after /synthesizing-cc-bigpicture. Chain the two skills:

/synthesizing-cc-bigpicture all 7d bigpicture.md
/persisting-bigpicture-learnings bigpicture.md docs/learnings/cc-bigpicture/

Compound Learning

Other skills can read hub-path/latest.md for prior synthesis context:

• synthesizing-cc-bigpicture — Load previous synthesis for incremental update instead of cold start.
• ralph / PRD skills — Reference latest bigpicture for strategic context.
• compacting-context — Include bigpicture summary when compacting at phase boundaries.

Quality Check

• latest.md content matches the most recent archive entry
• Archive files are append-only — never overwrite, never delete
• Hub path directories created automatically
• Input file validated before any writes