paid-channel-prioritizer
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted inputs during the data collection phase without explicit sanitization or boundary markers.
- Ingestion points: Phase 0 intake fields including Product URL, ICP descriptions, and Competitor names.
- Boundary markers: Absent; instructions do not provide delimiters or warnings to treat these values as literal strings.
- Capability inventory: The skill uses
web_search, executes local Python scripts, and performs file writes. - Sanitization: Absent; no escaping or validation of input strings is mentioned before use in commands.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands and write files to the local system.
- Evidence: Phase 1B explicitly directs the agent to run
python3scripts for Meta and Google ad scraping. Phase 4 instructions involve writing report files to specific directory paths.
Audit Metadata