ai2human-task-router
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from human executors (proofs, notes, and screenshots) which could contain malicious instructions.
- Ingestion points: Data returned from the
statusUrlandtaskUrlendpoints (SKILL.md). - Boundary markers: The skill includes explicit instructions for the agent to treat all human-generated content as untrusted and avoid executing scripts or commands found in proof content.
- Capability inventory: The skill primarily uses network requests to the vendor API and requires Node.js for a local smoke test utility (scripts/smoke.mjs).
- Sanitization: Instructions mandate human-in-the-loop verification and independent validation of any links or instructions returned in human proof.
- [CREDENTIALS_UNSAFE]: The skill documents the use of an
AI2HUMAN_API_KEYbut follows industry best practices by instructing users to use secure secret managers and warning against exposing keys in task content or logs. - [COMMAND_EXECUTION]: A smoke test utility (
scripts/smoke.mjs) is provided for developers. It includes security checks that prevent the accidental transmission of API keys to non-local or untrusted URL overrides. - [EXTERNAL_DOWNLOADS]: The skill references official vendor domains (ai2human.io) and the author's GitHub repository (BankrBot/skills) for installation and API interactions, which are legitimate service endpoints.
Audit Metadata