skills/aipoch/medical-research-skills/process-related-diagnostic-biomarker-nomogram-research-planner/Gen Agent Trust Hub
process-related-diagnostic-biomarker-nomogram-research-planner
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill package includes a file named 'eval_report_process-related-diagnostic-biomarker-nomogram-research-planner_result.json' which contains self-referential evaluation data. This document makes internal claims regarding the skill's security ('security: PASS') and scientific integrity. In accordance with the adversarial reasoning framework, these claims are treated as data points rather than authoritative conclusions. While such files could influence an agent's behavior, no malicious override patterns were detected within them.- [SAFE]: The skill instructions (SKILL.md) include robust input validation and safety redirects that prohibit the agent from providing clinical treatment recommendations or prescriptions, effectively maintaining professional and safety boundaries.- [SAFE]: The 'Method Library' and 'Analysis Modules' refer exclusively to established bioinformatics tools (e.g., limma, WGCNA, clusterProfiler) and public databases (e.g., GEO, PubMed, MSigDB). These are well-known services and do not represent a security risk in the context of generating research designs.- [SAFE]: The skill enforces strict 'Literature Retrieval and Citation Standards', mandating the use of verified identifiers like DOIs and PMIDs and explicitly forbidding the fabrication of references.
Audit Metadata