ecosystem-research
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection via the ingestion of untrusted external sources such as technical blogs, product documentation, and public precedents.
- Ingestion points: Data is gathered from external 'ecosystem' sources and project 'internal' sources (PRs, issues, repository history) as described in SKILL.md.
- Boundary markers: The RESEARCH-RIGOR.md file mandates the separation of direct evidence, interpretation, and applicability limits.
- Capability inventory: The skill spawns subagents to perform research, writes files to the OS temporary directory, and accesses project repository history.
- Sanitization: A completion gate in RESEARCH-RIGOR.md requires validation of report quality, source link presence, and evidence-versus-inference separation before synthesis.
Audit Metadata