skills/marcusrbrown/systematic/proof/Gen Agent Trust Hub

proof

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute curl commands for all API and bridge interactions.
  • [EXTERNAL_DOWNLOADS]: Fetches document state and configuration from www.proofeditor.ai and the local host bridge at localhost:9847.
  • [DATA_EXFILTRATION]: Transmits document markdown and metadata to the external service proofeditor.ai, which is essential for its primary document-sharing functionality.
  • [PROMPT_INJECTION]: Reads and processes untrusted markdown content from external documents, creating a potential surface for indirect prompt injection.
  • Ingestion points: Document content retrieved from API endpoints.
  • Boundary markers: Not defined in the instructions.
  • Capability inventory: Tools like Bash, Write, and WebFetch are available to the agent.
  • Sanitization: No content validation or sanitization of external document data is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 01:22 PM
Security Audit — agent-trust-hub — proof