Coordinate multi-agent codebase evaluations and generate remediation reports.
npx skills add https://github.com/HatmanStack/news-investor --skill repo-eval
Please help me install this Agent Skill. Skill: repo-eval Source: https://github.com/HatmanStack/news-investor/tree/main/.claude/skills/repo-eval Command: npx skills add https://github.com/HatmanStack/news-investor --skill repo-eval
This Skill automates the comprehensive evaluation of a codebase against defined role levels, identifying areas for improvement and remediation.
eval.md) outlining scores, remediation targets, and detailed feedback from each agent.Use the repo-eval skill to evaluate the current codebase for a Mid-Level Developer role.
High-intent search queries and answers about installing and using this skill.
You can evaluate a codebase by scoping the assessment against a defined role level, such as a Senior Developer benchmark, to generate a detailed remediation report on code quality and architecture. The system coordinates specialized agents to assess the project from different perspectives.
Assessing a take-home project objectively involves running parallel specialized evaluator agents that analyze code quality, architecture, and maintainability against a customized benchmark. This generates a consolidated document with scores and detailed feedback from multiple engineering perspectives.
Yes, code review evaluations can be customized by gathering user input on specific focus areas and concerns before the assessment begins. This scoping process tailors the multi-agent evaluation to target the exact architectural or maintainability issues you need to analyze.
Multi-agent code evaluation works by deploying three distinct AI personas—a Pragmatist, an Oncall Engineer, and a Team Lead—to assess a codebase in parallel. Their individual findings are consolidated into a structured markdown document with scores and remediation targets.
The codebase evaluation report is generated as a structured markdown document named eval.md. This file consolidates the findings from all parallel evaluator agents, providing detailed scores, remediation targets, and specific feedback for hiring or maintenance purposes.
muhammadhamza718Generate TypeScript validation logic for React forms with real-time error handling.
li3pIdentify and prioritize technical debt in a codebase for remediation planning.
0xRayAIAnalyze source code metrics, patterns, and dependency graphs.
nth5693Automate multi-perspective code reviews with structured checklists and workflow prompts.