Skills
skills/aradotso/mcp-skills/logos-distributed-reasoning-router/Gen Agent Trust Hub

logos-distributed-reasoning-router

Fail

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The installation section instructs the user to execute git clone https://github.com/rak7777/mythic-mcp-proxy.git followed by pip install -e .. This execution path allows arbitrary code from an unverified third-party repository to run on the user's system.
  • [EXTERNAL_DOWNLOADS]: The skill fetches code from a repository (github.com/rak7777/mythic-mcp-proxy.git) that is not associated with the skill author (Aradotso) or any established trusted provider.
  • [PROMPT_INJECTION]: The skill acts as a reasoning gateway that processes untrusted external data (such as PDF files and user prompts) without implementing visible boundary markers or sanitization logic, creating a surface for indirect prompt injection.
  • Ingestion points: router.query prompt parameter, workflow.add_step input_file parameter
  • Boundary markers: Absent
  • Capability inventory: Shell command execution via logos CLI, network requests via requests and websocket libraries
  • Sanitization: Absent
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 3, 2026, 04:48 AM
Security Audit — agent-trust-hub — logos-distributed-reasoning-router