The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a bash script scripts/scan.sh and multiple ripgrep (rg) commands in SKILL.md intended for searching patterns within a project's convex/ directory. These tools are used for diagnostic purposes to identify authentication and authorization gaps in the codebase.
[INDIRECT_PROMPT_INJECTION]: The skill and its included script ingest and process content from local files in the convex/ directory to identify security patterns.
Ingestion points: The scripts/scan.sh script and provided rg commands read source code files within the user's convex/ directory.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when processing file content during scanning.
Capability inventory: The skill's scripts use standard search tools (ripgrep) and output commands. There are no network write or file system write operations.
Sanitization: No sanitization is performed on the content read from files before it is processed by regex patterns, though the tool's impact is limited to reporting findings.

security-convex