triage-validation
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection by instructing the agent to ingest and analyze external, untrusted content from sources like HackerOne, Bugcrowd, and GitHub. The skill employs structured validation templates (the 7-Question Gate) as functional boundary markers for this data. Capabilities include security analysis and the use of illustrative HTTP request templates. Specific data sanitization routines are not present in this file.
- [SAFE]: The skill is a set of best practices and instructions for security researchers. It does not contain executable code, obfuscation, or persistence mechanisms. The "STOP" instructions are explicitly scoped to individual finding validation and do not interfere with global agent operation.
Audit Metadata