ce-work

Execute plan documents or work prompts through phased implementation, testing, and shipping workflows.

Updated Mar 17, 2026
One-click install
npx skills add https://github.com/Norfolk-Group/marcela-norfolk-ai --skill ce-work-norfolk-group
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ce-work
Source: https://github.com/Norfolk-Group/marcela-norfolk-ai/tree/main/skills/compound-engineering/skills/ce-work
Command: npx skills add https://github.com/Norfolk-Group/marcela-norfolk-ai --skill ce-work-norfolk-group

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning a plan or feature description into shipped code often breaks down through missed requirements, untested changes, and unfinished features. This Skill provides a structured execution workflow that takes a plan document or bare prompt and drives it through triage, environment setup, task tracking, incremental commits, quality checks, and PR creation. ## Core Features & Use Cases - Phased Execution Workflow: Triage input complexity, set up branches or worktrees, build task lists, and execute tasks with continuous testing and incremental conventional commits. - Subagent Dispatch: Run implementation units serially or in parallel with worktree isolation, file-overlap safety checks, and merge-conflict recovery procedures. - Quality and Shipping Gates: Enforce tiered code review, test coverage checks, linting, post-deploy monitoring plans, and PR creation with evidence and residual-finding tracking. - Use Case: A developer has a plan document with implementation units for a new API feature. Invoke this Skill with the plan path to have each unit implemented, tested, committed incrementally, reviewed, and shipped as a pull request. ## Quick Start Ask the agent to execute the plan at docs/plans/my-feature.md using the ce-work skill, or describe the work you want implemented directly.

Frequently Asked Questions about ce-work

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute a plan document with an AI coding agent?

Pass the plan file path as the argument to ce-work. The skill reads the plan's implementation units, builds a task list, sets up a branch or worktree, and executes each unit with testing and incremental commits before shipping a PR.

How does parallel subagent execution avoid file conflicts?

A Parallel Safety Check maps every unit's declared files and detects overlap before dispatch. With worktree isolation, overlaps surface as merge conflicts handled by serial re-dispatch; without it, the skill downgrades to serial subagents to prevent write races.

Can I use ce-work without a plan document?

Yes, a bare prompt describing the work is accepted. The skill scans affected files, assesses complexity, and routes trivial changes directly to implementation while larger scopes are routed to planning skills first.

When does ce-work escalate to deep code review?

Tier 2 review triggers when the diff touches sensitive surfaces like auth, payments, or migrations, exceeds 400 changed lines across multiple directories, exceeds 1000 lines total, or when the plan explicitly requests a thorough review.

What happens if tests fail during execution?

Failures are diagnosed and fixed immediately before proceeding. Dependent subagent units are never dispatched on a broken tree, and incremental commits are only created when tests pass.