OpenParable
Skip to content
← Back to skills

Done

Productivity
done

Done is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Completion marker for the /do workflow. Outputs hierarchical execution summary showing Global Invariants respected and all Deliverables completed. Called by /verify after all criteria pass, not directly.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/workflow/done-doodledood-claude-code-plugins-2/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/done, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

/done - Completion Marker

Goal

Output a completion summary showing what was accomplished, organized by the Manifest hierarchy.

Input

$ARGUMENTS = completion context (optional)

What to Do

Read the execution log and manifest. Output a summary that shows:

  1. Intent - What was the goal
  2. Global Invariants - All respected
  3. Deliverables - Each with its ACs, all passing
  4. Key changes - Files modified, commits made
  5. Tradeoffs applied - How preferences were used

Output Format

## Execution Complete

All global invariants pass. All acceptance criteria verified.

### Intent
**Goal:** [from manifest]

### Global Invariants
| ID | Description | Status |
|----|-------------|--------|
| INV-G1 | ... | PASS |

### Deliverables

#### Deliverable 1: [Name]
| ID | Description | Status |
|----|-------------|--------|
| AC-1.1 | ... | PASS |

**Key Changes:**
- [file] - [what changed]

---

### Tradeoffs Applied
| Decision | Preference | Outcome |
|----------|------------|---------|

### Files Modified
| File | Changes |
|------|---------|

---
Manifest execution verified complete.

Principles

  1. Mirror manifest structure - Hierarchy should match: Intent → Global Invariants → Deliverables
  2. Show evidence - Link changes to deliverables, describe behavioral changes (not just file lists)
  3. Adapt detail to complexity - Simple task = condensed output. Complex task = full hierarchy.
  4. Called by /verify only - /done is the final step after /verify confirms all criteria pass. If the execution log doesn't show verification, something went wrong upstream.

Medium Routing

When medium is non-local, /done output goes through the calling context (/verify → /do). No special routing needed — just produce the summary as normal.