How do I trigger Coco Execute?

Run the /coco:execute command; it handles the full 15-step TDD workflow with PRs and AI review.

What happens if pr.enabled is false?

The flow runs without creating or merging PRs, but still performs TDD, pre-commit checks, and tracker updates.

How many AI review iterations are allowed?

Up to 3 review-fix iterations are performed by the code-reviewer agent.

execute

npx machina-cli add skill skullninja/coco-workflow/execute --openclaw

Files (1)

SKILL.md

1.6 KB

Coco Execute Skill

This skill delegates to the /coco:execute command, which implements the full 15-step TDD execution loop with PR workflow, AI code review, and issue tracker bridge sync.

When to Use

Implementing a feature that has been imported into the coco tracker as an epic
Executing tasks from a specs/{feature}/tasks.md that has been converted to tracked tasks
Any multi-session work where dependency tracking matters

Alternatives

For single-issue hotfixes or quick changes, use the hotfix skill instead
For autonomous execution until epic completion, use /coco:loop

Execution

Run the /coco:execute command. It handles:

Pre-execution gate (tracker + issue tracker verification)
Dependency-aware task selection via bash "${CLAUDE_PLUGIN_ROOT}/lib/tracker.sh" ready
Issue branch creation (if pr.enabled)
Issue tracker bridge (start: "In Progress")
TDD implementation (RED -> GREEN -> verify)
Pre-commit validation (UI change detection)
Commit with issue key traceability
PR creation with issue ID (if pr.enabled)
AI code review via code-reviewer agent
Review-fix loop (max 3 iterations)
PR merge (if pr.enabled)
Tracker task close
Issue tracker bridge (complete: "Done" at PR merge)
Acceptance criteria verification
Next task check (loop or report completion)

Source

git clone https://github.com/skullninja/coco-workflow/blob/main/skills/execute/SKILL.mdView on GitHub

Overview

Executes the next available tracked task using a structured 15-step TDD loop. It orchestrates PR workflow, AI code review, and issue-tracker bridge sync to support cohesive, multi-session feature work.

How This Skill Works

The skill delegates to /coco:execute, running a 15-step loop that includes pre-execution gate checks, dependency-aware task selection, optional issue branch creation, In Progress bridge updates, TDD (RED-GREEN-VERIFY), pre-commit validation, traceable commits, PR creation, AI code review with a 3-iteration fix loop, PR merge, tracker close, bridge completion, acceptance criteria verification, and moving to the next task.

When to Use It

Implementing a feature imported into the coco tracker as an epic
Executing tasks from specs/{feature}/tasks.md converted to tracked tasks
Coordinating multi-session feature work where dependency tracking matters
When PR workflow and AI code review are required
When performing autonomous execution until epic completion

Quick Start

Step 1: Run /coco:execute
Step 2: Confirm pre-execution gate and dependencies
Step 3: Allow the 15-step loop to complete and proceed to the next task

Best Practices

Ensure pr.enabled is set when PRs are required
Use dependency-aware selection via tracker.sh ready
Keep issue keys traceable in every commit
Leverage the AI code-reviewer for iterative fixes (up to 3 iterations)
Verify acceptance criteria before closing the task

Example Use Cases

Progressing feature X from epic to In Progress and closing on merge
Converting specs/{feature}/tasks.md into tracked tasks and executing sequentially
Coordinating multi-session work with dependent tasks across epics
Performing a non-PR hotfix flow when appropriate
Running continuous execution until epic completion with /coco:loop as fallback

Frequently Asked Questions

Add this skill to your agents