Audits, fixes, and commits each component of the Interleave app.
npx skills add https://github.com/antoinefink/interleave --skill deep-review-antoinefink
Please help me install this Agent Skill. Skill: deep-review Source: https://github.com/antoinefink/interleave/tree/main/.agents/skills/deep-review Command: npx skills add https://github.com/antoinefink/interleave --skill deep-review-antoinefink
Deeply review and harden the whole Interleave app, component by component, via a dynamic Workflow with one phase per part of the application (many parts — take your time). For each component an independent auditor verifies it is implemented as expected, the design works (light + dark, matches the kit), nothing is broken / weird / nonsensical, and there are enough tests — then a fixer loop (≤6 rounds) fixes EVERYTHING it finds (small wins and small mistakes included) unless a fix needs a massive refactor, and commits per component. Use when asked to "review the app in depth", "harden / audit everything", "do a full quality pass", or check the whole application is solid.
The skill decomposes the app into many components and runs an adversarial audit → fix → re-audit loop per component to surface defects, enforce design invariants, and ensure test coverage. It provides commit per component to preserve traceability and supports up to six rounds to reach a clean state. It relies on the Workflow tool to orchestrate cross-cutting checks (design fidelity, data integrity, and architecture invariants) across the full system.
Run a per-component deep review by initiating audits, applying fixes, and committing changes for every part of the Interleave app.
High-intent search queries and answers about installing and using this skill.
A per-component audit decomposes your app into parts and runs an adversarial audit, fix, and re-audit loop per component. It verifies implementation, design fidelity, and tests, then commits fixes per component to preserve traceability.
Harden application quality by using a dynamic workflow with up to six rounds per component. An independent auditor verifies implementation and design, then a fixer loop resolves all small wins and mistakes, committing changes to maintain design invariants.
Yes, a single quality pass can fix design fidelity issues and ensure test coverage. The workflow orchestrates cross-cutting checks for light and dark modes, matches the design kit, and verifies tests before committing the fixes per component.
Yes, the component audit workflow commits fixes automatically. It commits changes per component after the fixer loop resolves defects, which preserves traceability and maintains design invariants across the application.
The iterative component audit and fix loop runs for up to six rounds per component and fixes everything it finds unless a fix requires a massive refactor. It focuses on small wins and mistakes while maintaining overall architecture.
DYAI2025Identify software components from design artifacts and create per-component directories.
evanfang0054Test web applications for bugs and resolve them with regression testing.
ImaginerLabsGenerate prioritized E2E, API, component, and unit tests with fixtures and coverage summaries.
baphledGuide static code analysis tools to detect bugs, vulnerabilities, and style violations.
