The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes npx tsc --noEmit to verify the integrity and type safety of authorization fixes applied to the project code.- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes local source code from the convex/ directory (Ingestion points). The risk is mitigated by requiring objective regex-based matches to trigger actions (Boundary markers). The skill has the capability to modify project files and execute the TypeScript compiler (Capability inventory), and it uses tsc to verify the validity of generated code (Sanitization).- [SAFE]: The skill implements a 'Mandatory First Step' to verify the application's authentication foundation before applying authorization hardening, ensuring that enforcement is functional and correctly mapped to the user identity provider.

convex-authz