Identify and verify implicit code assumptions with tiered AI verification.
npx skills add https://github.com/williaby/.claude --skill rad
Please help me install this Agent Skill. Skill: rad Source: https://github.com/williaby/.claude/tree/main/skills/rad Command: npx skills add https://github.com/williaby/.claude --skill rad
This Skill systematically identifies and verifies implicit assumptions in your code, especially AI-generated code, that could lead to production failures. It prevents costly outages and data loss by catching these hidden issues before deployment, giving you peace of mind and more robust applications.
/rad/verify --scope=changed-files to automatically scan your new code for assumptions, verify them with AI, and receive suggested fixes to prevent potential bugs or outages.Verify all assumptions in your currently changed files using a tiered AI strategy.
/rad/verify --scope=changed-files
High-intent search queries and answers about installing and using this skill.
Identify assumptions by scanning code for implicit dependencies on timing, external resources, data integrity, and edge cases. RAD uses standardized tagging to flag critical, standard, and edge-case assumptions, then routes them to AI models for verification to catch issues before production.
Yes. Run `/rad/verify --scope=changed-files` to automatically scan new code, verify assumptions with tiered AI models based on risk level, and receive defensive code patterns and error handling suggestions to mitigate identified risks.
RAD checks assumptions across timing, external resources, data integrity, concurrency, security, and edge conditions. It works in code generation, code review, and CI/testing pipelines to prevent production failures from hidden implicit dependencies.
Risk-tier routing directs critical assumptions to premium AI models for thorough verification while routing standard and edge-case assumptions to free models. This optimizes verification depth and cost based on the assumption's impact on production stability.
Yes. RAD integrates with git, find, grep, and MCP dependencies. It fits into CI/testing pipelines and code review workflows, scanning changed files and generating remediation fixes compatible with standard development environments.
RAD generates verified assumption reports, defensive code patterns, and automated fix suggestions. These outputs guide remediation and help prevent bugs, data loss, and outages by showing exactly which assumptions need defensive handling.
ConaryLabsAnalyze git changes to classify modifications and assess impact and risk.
timurgaleevAutomate end-to-end QA testing and atomic bug fixes for web applications.
m19803261706Investigate, fix, test, and commit code changes with conventional commits.
huyhung9630Verify software changes with static analysis, unit tests, integration tests, and manual checks.