competitive-intelligence
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill manifest grants access to the
Bashtool. While intended for internal operations within the Product Org OS framework, the ability to execute shell commands provides a high-privilege execution environment. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) due to its data processing workflows.
- Ingestion points: The skill uses
WebSearchandReadtools to aggregate competitor data and market trends from arbitrary external websites (SKILL.md). - Boundary markers: The instructions lack specific delimiters or system-level directives to isolate external content or ignore instructions embedded within retrieved data.
- Capability inventory: The agent has access to
Bash,Write,Edit, andTasktools, which could be leveraged if the agent inadvertently obeys malicious instructions found in web content (SKILL.md). - Sanitization: There are no defined validation or sanitization routines to filter untrusted content before it is processed by the model.
Audit Metadata