waypoint-bio
Fail
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [REMOTE_CODE_EXECUTION]: Automated scanners detected a malicious URL (https://www.outpost.bio/citations) within the skill documentation, which is flagged on security blacklists.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted microbiome taxonomic data from various profiler outputs without implementing prompt boundary markers. Ingestion points: Data preparation scripts and CLI subcommands like prepare-dataset. Capability inventory: The skill can read/write local files and perform network operations to Hugging Face. Sanitization: Employs basic taxonomic string normalization in scripts/profiler_to_waypoint.py.
- [DYNAMIC_EXECUTION]: The skill loads a tokenizer from Hugging Face with the trust_remote_code parameter enabled, which executes remote Python code during model initialization.
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the waypoint-bio package from PyPI and fetches models, datasets, and benchmarks from gated Hugging Face repositories.
Recommendations
- AI detected serious security threats
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata