prfaq-beagle

Pass

Audited by Gen Agent Trust Hub on Apr 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill follows intended workflows for interaction and state management within its own workspace.
  • [COMMAND_EXECUTION]: The skill utilizes platform-native tool calling to invoke companion skills for data grounding. This is an orchestrational pattern for multi-skill workflows.
  • [DATA_EXFILTRATION]: File operations are restricted to the designated workspace directory and local project documentation. No unauthorized network activity or sensitive file access was observed.
  • [PROMPT_INJECTION]: The skill processes external data from local documents and web research.
  • Ingestion points: Reads report files in .beagle/concepts//analysis/ and research/ via the artifact-analysis and web-research companions.
  • Boundary markers: Uses Markdown headers and explicit Reasoning blocks to structure findings.
  • Capability inventory: Limited to file-writing within its workspace and specific tool calls; lacks arbitrary shell or code execution.
  • Sanitization: Relies on the agent's internal filtering for external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 19, 2026, 05:59 AM