# Run Markdown Implementation Tasks

Implementation plans often stall when tasks are tracked outside the design document. This skill reads Markdown checklists, selects the next task, and updates progress in place.

## Install

```bash
npx skillstore add notedit/task-execution-engine
```

## Metadata

- Status: approved
- Slug: notedit-task-execution-engine
- Skillstore revision: r1
- Version status: missing
- Tree hash: 8c39dc7d5c4fa69157328b8d8436bb6e121116697aa732ed4c079e23acd7f4d3
- Author: notedit
- GitHub username: notedit
- License: MIT
- Repository: https://github.com/notedit/happy-coding-agent/tree/main/skills/task-execution-engine
- Ref: 62e2a730c5cd74eab4c7164309d810de660fcea3
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 72
- Public page: https://skillstore.pages.dev/skills/notedit-task-execution-engine
- Manifest: https://skillstore.pages.dev/api/skills/notedit-task-execution-engine/manifest

## Capabilities

- Parses Implementation Tasks sections from Markdown files.
- Selects the next pending task by priority after dependencies are complete.
- Marks tasks completed, failed, or pending with checkbox updates.
- Summarizes completed, pending, blocked, and failed task counts.
- Outputs task details as text or JSON through the helper script.

## Use Cases

- Resume a Feature Build: Continue from the first unfinished task in a design document after an interrupted coding session.
- Coordinate Batch Implementation: Run a prioritized task list while tracking completed, blocked, and failed work in one Markdown file.
- Track Design-Driven Work: Keep implementation status beside requirements, target files, and acceptance criteria.

## Prompt Templates

### Start the Next Task

```
Use task-execution-engine with docs/designs/user-auth.md. Show the next available task before editing files.
```

### Resume Interrupted Work

```
Resume implementation from docs/designs/user-auth.md. Continue from the first pending task and update checkbox status after each result.
```

### Run With Verification

```
Run tasks from docs/designs/billing.md in dependency order. After each task, run relevant tests and mark failures with a short reason.
```

### Audit Then Continue

```
Review docs/designs/search.md, summarize completed, pending, blocked, and failed work, then implement the highest-priority unblocked task.
```

## Limitations

- It only understands the documented Markdown checkbox format.
- It does not validate whether implementation code is correct.
- It can rewrite any file path provided to the helper script.
- It does not create backups before updating task files.

## Best Practices

- Review the design document before starting unattended execution.
- Keep task titles stable because dependencies reference task titles.
- Create a checkpoint before running commands that update task files.

## Anti Patterns

- Do not use vague tasks without files or acceptance criteria.
- Do not run it against arbitrary files outside the project plan.
- Do not mark a task complete before verification.

## Security Audit

- Audited at: 2026-07-05T19:58:28.106\+00:00
- Summary: Most static findings are false positives caused by Markdown code fences, inline task attributes, and regex patterns. No shell or Ruby backtick execution was found. One medium issue remains because the helper reads and rewrites any supplied file path without validation.

## Stats

- Views: 189
- Downloads: 15
- Favorites: 0
- Popularity score: 0