Coordinate specialist reviewers to audit codebases and generate consolidated Markdown reports.
npx skills add https://github.com/howells/arc --skill audit-howells
Please help me install this Agent Skill. Skill: audit Source: https://github.com/howells/arc/tree/main/skills/audit Command: npx skills add https://github.com/howells/arc --skill audit-howells
This Skill orchestrates comprehensive codebase audits by coordinating a panel of specialist reviewers. It yields actionable reports that help you uncover security gaps, performance bottlenecks, architectural concerns, and quality issues before shipping.
Run the audit on a chosen scope (for example, /arc:audit apps/api) to generate a unified report of findings.
High-intent search queries and answers about installing and using this skill.
To run a codebase audit, execute the audit command on a chosen scope. It coordinates multiple specialist reviewers in batches by default to generate a single consolidated markdown report of security, performance, and architecture findings.
Yes, codebase audits support scope flexibility by allowing you to target a specific path like apps/web instead of the entire project. This focuses the reviewer orchestration and generates a consolidated report strictly for that directory.
To speed up batch processing during a code review, use the optional --parallel flag. This executes the panel of specialist reviewers simultaneously rather than in fixed-size batches, yielding the consolidated markdown report faster.
A multi-review codebase audit covers diverse expertise areas including security, performance, architecture, organization, UI, and data engineering. This holistic coverage uncovers gaps and bottlenecks before shipping your code.
No external dependencies are required to perform an architecture review. The audit process relies on a defined internal set of specialist reviewers, such as the architecture-engineer, to analyze the codebase and deliver actionable findings.
Generating a consolidated markdown report for a security review merges findings from all specialist reviewers into a single document. This streamlines remediation by providing a unified view of security gaps and architectural concerns.
jamie8johnsonCoordinate parallel agents across batches to audit codebases across 14 categories.
Snowflake-LabsA concise summary of your roleplay character: name,</name> personality, and/or behavior, formatted as a single string sentence. Typically不超过50字，此处省略
alexei-ledDetect languages in diffs and run linters, security checks, and coverage analysis.
koya-altScan codebases for dependency, static, and data-flow vulnerabilities.
