design-consultant-feedback
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data such as meeting notes, transcripts, and sprint recaps in Mode 1.
- Ingestion points: The instructions in
SKILL.mddirect the agent to "Read the content provided" and evaluate it against consulting best practices. - Boundary markers: No specific delimiters (e.g., XML tags or block quotes) or "ignore embedded instructions" warnings are defined for the user-provided text processing.
- Capability inventory: The skill is restricted to text generation and uses a structured interaction tool (
ask_user_input). It lacks high-risk capabilities such as network access, file system modifications, or arbitrary command execution. - Sanitization: There is no explicit instruction to sanitize, filter, or validate the user-provided content before it is processed by the model.
- [SAFE]: The skill is authored by thoughtbot and incorporates their established professional design guidelines. All external references point to official vendor domains (thoughtbot.com) or well-known services (GitHub). No patterns related to obfuscation, credential theft, or unauthorized persistence were detected.
Audit Metadata