skills
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external documents such as Terms of Reference (ToR) and Request for Proposals (RFP) into the agent's context. This presents a surface for indirect prompt injection if those documents contain malicious instructions. However, the skill includes references to quality gates and audits (e.g., meta/anti-ai-slop) to mitigate against low-quality or untrusted content.
- Ingestion points: SKILL.md (ToR, RFP, advert, proposal brief).
- Boundary markers: Mention of meta/anti-ai-slop and ai-slop-audit as gates.
- Capability inventory: SKILL.md (Read, search, edit).
- Sanitization: Described in methodology as "abstain logic" and "hallucination ceilings," though not implemented via executable code in the provided files.
- [NO_CODE]: The skill consists entirely of Markdown instructions, templates, and reference materials. No executable scripts (Python, JavaScript, shell) or configuration files that trigger code execution were identified in the provided content.
- [SAFE]: No specific security threats such as credential exfiltration, persistence mechanisms, or unauthorized command execution were detected in the analyzed files. The skill appears to be a legitimate collection of professional templates and guidance for consulting services.
Audit Metadata