linkedin-company-page
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Bash tool to run a local Python script for LinkedIn API interactions. The script is part of the skill package and performs specific, documented tasks.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests text from a file to publish it to a LinkedIn feed, creating a surface where malicious instructions could potentially be embedded in the content being posted.\n
- Ingestion points: Reads the content of post.txt in scripts/linkedin_company_page.py to populate the LinkedIn post commentary.\n
- Boundary markers: The ingested text is handled as a data string within a JSON payload, providing structural separation from the API logic.\n
- Capability inventory: The script performs authenticated HTTP POST and GET requests to api.linkedin.com and uses the publish_artifact tool to record successful posts.\n
- Sanitization: The script validates that the organization ID is a positive numeric value and strips whitespace from the post content.
Audit Metadata