reactome-database

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

• Third-party content exposure detected (medium risk: 0.30). Runtime reads outsider-authored free text when a user supplies an identifier list or TSV expression data to Reactome’s AnalysisService via POST /AnalysisService/identifiers/ (including user-controlled gene_list_file content in scripts/reactome_query.py::command_analyze).

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I inspected the documentation and example code for literal credentials. The only literal credential-like value is the analysis token "MzUxODM3NTQzMDAwMDA1ODI4MA==", shown verbatim in the API reference examples. This is a concrete token (base64) that could be used with the GET /token/{token} endpoint to retrieve results, so it meets the definition of a secret (literal token value). No API keys, private key blocks, or other high-entropy credentials were found. Other values are endpoints, IDs, or obvious placeholders/examples and were ignored.