scp-instructproteinkg

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt embeds an API key header in config and code examples (e.g., headers: "SCP-HUB-API-KEY": "{API-KEY}" and transport headers={"SCP-HUB-API-KEY": "sk-xxx"}), which instructs placing the secret directly into generated config/code/requests and therefore requires the LLM to handle/output secret values verbatim.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). The required runtime path is calling the MCP tool get_experiment_workflow (and related KG tools) which returns JSON “workflow (steps + reagents)” originating from the SciGraph server’s stored experiment content; that content is not authored by the operating user and is therefore outsider-provided free text/fields that can be ingested into the agent’s LLM context.