3,748 tools found
A companion skill that turns a caught drift into a permanent WHEN/DON'T/INSTEAD project memory clause.
Read-only audit skill for reviewing agent run evidence without rerunning tools, approving actions, or modifying data unless explicitly authorized.
A structured analysis format for AI assistants that runs before vulnerability hunting, enabling ultra-granular function-level context extraction.
Skill that statically audits AI skills and bundles for malicious patterns, data leaks and obfuscated payloads across desktop and mobile platforms.
Bound ineffective coding repairs with failure fingerprints, a three-attempt limit, real-path checks, negative controls, and tested rollback.
Audit authorized web apps for browser-visible secrets and exposed configuration using bounded public reads, redacted evidence, and contextual triage.
Skill with a six-step procedure to resume interrupted work by checking saved handoffs against the current repository and Git status.
Read-only RBAC audit skill that reviews marketplace role-based access control configurations without granting permissions to scan live services or alter data.
Turn Markdown tutorials, reports, or presentations into narrated MP4 videos with synchronized Marp slides, narration, and separate output files.
A skill for researching and citing sources via Parallel's free, anonymous Search MCP tools.
Skill for a bounded, evidence-based code review of a pinned diff from correctness, authorization and performance perspectives.
Skill that certifies a deployed service with fresh evidence across source, CI, migrations, readiness and critical routes into a GREEN, RED or BLOCKED verdict.
