alterlab-adaptyv

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes protein sequences and target identifiers provided by the user or fetched from the Adaptyv API. While it includes local validation steps (e.g., amino acid checking) to prevent common errors, there is a theoretical surface for indirect injection if malicious instructions were embedded in sequence labels or metadata retrieved from the platform.
  • [SECRET_MANAGEMENT]: The skill correctly advises users to store sensitive API keys in environment variables or gitignored .env files, which is standard secure development practice.
  • [EXTERNAL_SERVICES]: The skill interacts with legitimate biotech and academic services including the Adaptyv Bio Foundry API, DTU Health Tech's NetSolP, and Loschmidt Laboratories' SoluProt. These are established platforms in the biological research community.
  • [DEPENDENCIES]: The skill uses standard, widely-vetted Python packages such as requests, torch, pandas, and biopython. It also references official SDKs and research tools hosted on GitHub (e.g., ProteinMPNN), which is a common and accepted practice for bioinformatics workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 02:35 PM