setup-harmony-remote-tool
Fail
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to install a Python wheel file directly from a remote server at 'https://connect.aiscientist.tools/downloads/tuplatform_connect-0.3.0-py3-none-any.whl'. This domain is not identified as a trusted vendor or well-known service.
- [REMOTE_CODE_EXECUTION]: The installation of the external wheel file ('tuplatform-connect') via pip constitutes the execution of remote code on the host system. While the instruction includes a SHA-256 integrity hash ('sha256=3fad5eee5ecf7887a693d93ccd1aa112dc0955617a885d1fc3daded0030f9ae0'), the provenance of the code remains unverified by standard registries.
- [COMMAND_EXECUTION]: The skill requires the execution of multiple system-level commands, including virtual environment setup, pip installations, and the use of the 'tu' CLI for platform authentication ('tu remote login') and service deployment ('tu serve', 'tu remote share').
- [INDIRECT_PROMPT_INJECTION]: The implementation creates an attack surface for indirect prompt injection through data processing.
- Ingestion points: The 'run_harmony_integrate' tool processes external data files via the 'adata_path' parameter specified in 'SKILL.md'.
- Boundary markers: No specific delimiters or safety instructions are defined to separate tool parameters from user-provided data.
- Capability inventory: The environment possesses broad capabilities including shell execution ('tu', 'python'), file system modification ('mkdir', 'rm'), and network relaying.
- Sanitization: The setup instructions do not document any sanitization, schema validation, or content filtering for the processed data files.
Recommendations
- AI detected serious security threats
Audit Metadata