pitch-recon
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests untrusted data from the local repository and external websites without implementing boundary markers.
- Ingestion points: The skill searches and reads various file types (e.g., .md, .html, .tsx) and fetches external content via the WebFetch tool.
- Boundary markers: Instructions do not include delimiters or warnings to ignore instructions that might be present in the content being analyzed.
- Capability inventory: The skill utilizes the Bash tool for file discovery and WebFetch for retrieving network data.
- Sanitization: No content sanitization or validation processes are defined.
- [COMMAND_EXECUTION]: The skill uses shell commands through the Bash tool (e.g., find, grep, xargs) to discover files containing marketing-related keywords within the repository.
- [EXTERNAL_DOWNLOADS]: The skill is configured to retrieve data from external URLs and web searches using the WebFetch and WebSearch tools to gather marketing intelligence.
Audit Metadata