qbr-prep
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as CRM exports, meeting transcripts, and uploaded PDFs, creating a surface for indirect prompt injection.
- Ingestion points: SKILL.md (Step 1) and references/input-rules.md (Rule 1, Rule 5) define the ingestion of external content.
- Boundary markers: SKILL.md (Section 0) contains a mandatory 'Untrusted content' rule instructing the agent to treat external text as data and ignore any embedded instructions.
- Capability inventory: The skill generates narrative text and has the capability to produce PPTX files as described in the 'QBR Deck Option' section of SKILL.md.
- Sanitization: The skill relies on policy-based instructions to prevent the execution of embedded commands found in user-provided data.
Audit Metadata