social-selling
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to research and ingest data from LinkedIn profiles, posts, and comments to personalize outreach.
- Ingestion points: Prospect LinkedIn activity, posts, and comments (defined in
SKILL.md, Phase 2). - Boundary markers: The instructions do not specify boundary markers or delimiters for the ingested social media content.
- Capability inventory: The skill provides a local Python script
scripts/check-output.pywhich performs file read operations. It relies on the agent's built-in tools for web browsing and research. - Sanitization: No specific sanitization or validation of the external content is mentioned.
- [COMMAND_EXECUTION]: The skill includes a Python script
scripts/check-output.pyintended to be run locally to validate that the social selling deliverables meet minimum keyword requirements. The script uses standard libraries and performs simple string checks on local files.
Audit Metadata