design-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from users and propagates it through an automated pipeline.
- Ingestion points: User-provided research findings or problem descriptions are collected in
SKILL.md(Stage 1) and used to create a 'Research Report' that serves as context for all subsequent stages. - Boundary markers: The skill lacks explicit instructions or delimiters to prevent the agent from following instructions that might be embedded within the user-provided research text.
- Capability inventory: The skill performs automated file writes (HTML reports) and invokes multiple sub-skills (
user-researcher,product-designer,accessibility-auditor, etc.) that modify project files. - Sanitization: There is no mention of sanitizing or validating the user input before it is used to drive the automated design and auditing stages.
- [INDIRECT_PROMPT_INJECTION]: The 'Automated Mode' feature instructs the agent to run the entire 9-stage pipeline without pausing for user confirmation. This reduces human oversight, increasing the risk that malicious instructions hidden in the input data could influence the final design outputs or generated code/HTML without detection.
Audit Metadata