competitor-product-monitor
Warn
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands that incorporate user-supplied or inferred input without adequate sanitization. In
SKILL.md(Step 3), the instructiontinyfish agent run --sync --url "<competitor_url>"allows for command injection if the<competitor_url>contains shell metacharacters (e.g.,"; malicious_command #). - [INDIRECT_PROMPT_INJECTION]: The skill crawls and processes content from external third-party websites, which are untrusted sources.
- Ingestion points: External competitor URLs including blogs, changelogs, and newsrooms as defined in
SKILL.md(Step 3). - Boundary markers: Absent. The instructions do not define delimiters or warnings for the agent to ignore instructions that might be embedded in the fetched HTML or markdown.
- Capability inventory: Execution of shell commands via the TinyFish CLI and report generation.
- Sanitization: Absent. The skill expects structured JSON from the web agent but does not verify or sanitize the content before presenting it in the final report.
Audit Metadata