Revive

Inspired by the REVIVE archetype from the /goal video. Walks every project in the vault, follows the frontmatter path: field to its on-disk location, checks git activity + remaining tests + last-edited dates, and surfaces dormant projects with concrete one-page resurrection plans.

When to use

Triggers: revive, resurrect, dormant projects, what projects could I pick back up, find stalled work, what have I been ignoring.

Best run quarterly or when Andy is between major sprints and wants to scan for next-up candidates.

Load

• 02 - Projects/ - all project pointer notes (each has path: frontmatter)
• Projects-Index.md - the auto-rolled list
• CLAUDE.md - drive map for context on where projects live
• For each project: the on-disk folder per path: field

Run

Phase 0 — Inventory

List every 02 - Projects/<category>/<project>.md with: status (from frontmatter), category, path on disk.

Phase 1 — Per-project scan

For each project:

• Read the pointer note
• Visit path: location on disk
• git log -1 for last commit date
• Count files in the project folder (rough scope signal)
• Check for README.md, ROADMAP.md, NEXT-SESSION-HANDOFF.md — these are resurrection material
• Look for tests / scripts that might still run

Phase 2 — Classify

Phase 3 — Resurrection plan (recently/cold dormant only)

For each dormant candidate, produce a one-paragraph plan:

• State at pause — what was the last thing being worked on (from last commit message + latest README/ROADMAP)
• First move to resume — the single most useful 30-minute action to get unstuck
• What's at risk if not revived — does this depend on something that's bit-rotting?
• Suggested model + skill — /goal if the resumption is iterative; direct work if it's a single step

Output

Write a revive report to 01 - Daily Notes/<YYYY-MM-DD> - Revive Report.md:

---
type: revive-report
created: <MM-DD-YY>
projects_scanned: <count>
active: <count>
recently_dormant: <count>
cold_dormant: <count>
archived: <count>
stubs: <count>
tags: [revive, projects]
---

# Revive Report -- <MM-DD-YY>

## Recommended pickups (ordered by ease + value)

### 1. <project-name>
- **Status:** recently dormant (last commit MM-DD-YY)
- **State at pause:** <one line>
- **First move:** <30-min action>
- **Risk:** <low/medium/high>
- **Suggested approach:** <model + skill>

### 2. <project-name>
...

## Cold dormant (revival possible but costly)
- <project>: <reason it might still matter>

## Stubs to resolve
- <project>: pointer note at <path> but on-disk path missing. Delete or scaffold?

## Leave alone (archived)
- <project>: <reason>

After

1. Surface the top 1-3 pickup candidates with their first moves
2. Don't auto-resume — Andy chooses what to pick up
3. If Andy picks a candidate, hand off to /project-scaffolder if scaffolding is needed, or just open the project's path
4. Append Activity Log row

Don't

• Don't propose reviving archived projects unless Andy explicitly asks
• Don't auto-delete stub pointer notes — surface them, let Andy decide
• Don't grade projects by line count — small-but-shipped beats big-but-stalled
• Don't ignore the path: frontmatter — projects live on disk, not in the vault
• Don't run on every project list at once if there are >20 — paginate the report