enhance-web-redesign
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent reads and modifies a project's codebase, creating a surface where malicious instructions embedded in processed code could influence agent behavior.
- Ingestion points: The agent is instructed to 'Scan' and 'Read the codebase' to identify framework and styling patterns (SKILL.md).
- Boundary markers: Absent. There are no explicit instructions to treat code content as untrusted data or to use delimiters to prevent instruction leakage.
- Capability inventory: The skill allows for file modification and styling upgrades ('Fix' phase in SKILL.md).
- Sanitization: No sanitization or validation of the ingested code content is described.
- [EXTERNAL_DOWNLOADS]: The skill references external websites and repositories for research and placeholder assets.
- Evidence: Links to design research repositories at
github.com/awaken7050dev/anti-slop-uiandgithub.com/Leonxlnx/taste-skill. - Evidence: Recommends fetching placeholder imagery from
picsum.photosduring the design audit phase.
Audit Metadata