Audit agent codebases against a 10-axis design rubric with prioritized issues.
npx skills add https://github.com/s0613/S-skills --skill sj-agent-review
Please help me install this Agent Skill. Skill: sj-agent-review Source: https://github.com/s0613/S-skills/tree/main/skills/sj-agent-review Command: npx skills add https://github.com/s0613/S-skills --skill sj-agent-review
Manually auditing business agent codebases for structural flaws, missing guardrails, and design inconsistencies is time-consuming and prone to human error, often letting critical issues slip through until they cause runtime failures or security vulnerabilities in production. This Skill automates the audit process using a standardized 10-axis design rubric, ensuring no key quality checks are missed before deployment.
Use the sj-agent-review skill to audit the agent project in the ./my-business-agent directory against all 10 design axes and receive a prioritized list of improvement actions.
High-intent search queries and answers about installing and using this skill.
Automated agent code review audits business agent codebases against a 10-axis design rubric to identify structural flaws, missing guardrails, and runtime risks. It generates per-axis scores, PASS/WARN/FAIL verdicts, and prioritized issue lists with file:line references and actionable fixes.
A structural audit for AI agent design quality scores agent projects across 10 core dimensions including orchestration, role separation, and guardrails. It identifies design inconsistencies and missing safety mechanisms with a 0-10 point scale, ensuring no key quality checks are missed before production deployment.
Yes, you can audit a single design axis like context management or guardrails instead of the full project. The structural audit supports focused reviews of individual design dimensions based on your request, making it useful for both full pre-deployment checks and targeted deep dives.
The agent design audit applies to AI agent systems built with any programming language or framework. It evaluates structural flaws and runtime risks using a standardized rubric, ensuring framework-agnostic quality assessment for pre-deployment checks and legacy codebase reviews.
The 10 design axes evaluated in an agent code quality check are runtime loop, orchestration, role separation, tool hierarchy, context management, guardrails, observability, memory layer, evaluation, and graph topology. Each axis receives a 0-10 score with PASS/WARN/FAIL verdicts.
A pre-deployment quality gate for business AI agents is needed because manual auditing is time-consuming and prone to human error, often letting critical runtime failures or security vulnerabilities slip through. Automating this process with a standardized rubric ensures consistent structural integrity and missing guardrail detection.
az9713Initialize a project with dark code prevention directories, CLAUDE.md section, and baseline audit.
lucasFR2Audit code for structural flaws, fragility, and production risks.
klu-devManages lifecycle, observability, and security controls for cloud-hosted agent systems.
habit-hooksDetects hidden correctness bugs, design flaws, and missing test coverage in change sets after linting passes.
