social-graph-ranker
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. It processes untrusted data from social media platforms to rank connections and draft outreach messages. An attacker could influence the agent by including malicious instructions in their public social media profile or bio. (1) Ingestion points: LinkedIn CSV file (~/Downloads/Connections.csv) and X API user data. (2) Boundary markers: None identified. (3) Capability inventory: Network access via X API, file read for CSV data, and message drafting. (4) Sanitization: No specific filtering or sanitization of external content is mentioned.
- [DATA_EXFILTRATION]: The skill accesses sensitive personally identifiable information (PII) by reading LinkedIn connection exports (~/Downloads/Connections.csv). While this is the intended purpose of the tool, users should be aware of the level of personal data access required.
- [EXTERNAL_DOWNLOADS]: The skill performs legitimate network requests to the X API and integrates with the Exa search service to gather graph data and lead intelligence.
Audit Metadata