Generates cryptographic test vectors by running mutation testing to find code paths that existing vectors don't exercise, then creates new vectors that deliberately hit those gaps. You run a baseline mutation test, analyze which mutants escaped, write vectors that kill them, then compare before and after kill rates to prove effectiveness. Works with any crypto primitive that has multiple implementations and a mutation testing framework. The workflow is rigorous: it requires both positive and negative test cases (rejecting invalid inputs, not just accepting valid ones), roundtrip assertions, and cross-implementation coverage. Best for Wycheproof-style test suites where you need quantifiable proof that your vectors actually improve coverage, not just intuition.
npx -y skills add trailofbits/skills --skill vector-forge --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
moizibnyousaf/ai-agent-skills
github/awesome-copilot