The Agent Skills Directory

[NO_CODE]: The skill consists solely of markdown instructions describing a triage process. It does not include any scripts, source code, or binary files.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from a findings.json file, which creates a potential vulnerability where malicious instructions embedded in the findings could be interpreted by the AI agent. Ingestion points: Reads findings.json to profile and summarize data. Boundary markers: The instructions lack specific delimiters or system-level directives to isolate external content from the prompt context. Capability inventory: The skill has no capabilities to execute code or perform network operations; it is limited to reasoning and local file creation. Sanitization: No validation or escaping of the input data is mentioned in the instructions.

github-triage-findings