comprehensive-enrichment

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill transmits user-provided parameters, such as email addresses, names, and LinkedIn URLs, to external service endpoints (e.g., *.mpp.paywithlocus.com and *.mpp.tempo.xyz) to perform data enrichment. This behavior is documented as the core functionality of the skill.
  • [PROMPT_INJECTION]: The skill ingests data from external sources via web search results (Exa) and web scraping (Firecrawl), which creates a surface for indirect prompt injection where content from web pages could influence the agent.
  • Ingestion points: manifest.json (steps involving Exa search, Exa findSimilar, and Firecrawl extract)
  • Boundary markers: No specific delimiters or instructions to ignore embedded content are defined in the manifest steps.
  • Capability inventory: The skill is configured to perform HTTP POST requests and return data; it does not define local filesystem access or shell command execution capabilities.
  • Sanitization: No explicit filtering or validation of the retrieved external content is defined within the skill's manifest logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 07:34 PM
Security Audit — agent-trust-hub — comprehensive-enrichment