implement-qc-loop

Solid

Runs Implement → QC in a continuous loop until QC passes or a safety limit is reached. Combines /sddp-implement and /sddp-qc into a single uninterrupted workflow.

AI & Automation 96 stars 9 forks Updated today MIT

Install

View on GitHub

Quality Score: 85/100

Stars 20%
66
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
80
License 10%
100
Description 5%
100

Skill Content

# Software Engineer — Implement + QC Loop Workflow <rules> - Orchestrates `/sddp-implement` + `/sddp-qc` in a single turn. Loads and executes each sub-skill inline — does not duplicate their logic. - Executes for real. Not a demo, dry run, or simulation. - Never treat marker creation alone as success. `.completed`/`.qc-passed` valid only when backed by actual work and report state. - Artifacts inconsistent with `tasks.md`, `qc-report.md`, or the `.qc-passed` evidence digests → halt and surface. - **Safety limit**: Max **10** iterations → halt with latest `qc-report.md`. - Report brief status at each iteration boundary: iteration number, bug tasks added, remaining failures. - Same gating rules as sub-skills: `spec.md`, `plan.md`, `tasks.md` required. - **NEVER yield between iterations** — continuous turn until QC passes or safety limit. - **Halt early** (yield to user) if: 1. Implement halts due to sequential task double-failure (autopilot: automatic halt; interactive: user chooses "Halt") 2. QC generates `manual-test.md` 3. Implement cannot produce `.completed` (catastrophic failure) 4. QC finds only CRITICAL `project-instructions.md` violations 5. Marker/report state inconsistent with actual evidence - **Artifact conventions**: All sub-skill rules in `AGENTS.md` §Artifact Conventions apply. Never reverse checkboxes, delete task lines, or modify IDs. - Pass through user confirmation requests from sub-skills. - Optional `PIPELINE_CONTEXT` input: when supplied by `/s...

Details

Author
attilaszasz
Repository
attilaszasz/sdd-pilot
Created
6 months ago
Last Updated
today
Language
JavaScript
License
MIT

Similar Skills

Semantically similar based on skill content — not just same category