audit-ui-and-save-files
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill audits external web applications, creating a surface where malicious content on a target site could influence agent behavior.
- Ingestion points: Content is ingested from arbitrary URLs via the
run-agent-browsertool during Phase 3 of the audit (SKILL.md). - Boundary markers: Finding reports are written to disk using a strict Markdown template defined in
references/audit-format-spec.mdto help delineate external data. - Capability inventory: The skill can write files to the project directory, dispatch subagents via the
Agenttool, and modify source code through fix subagents (SKILL.md, Phase 5). - Sanitization: No explicit content sanitization or instruction-filtering for website data is described in the provided scripts or instructions.
- [EXTERNAL_DOWNLOADS]: The skill's documentation directs users to install components from the author's GitHub repository (
github.com/yigitkonur/skills-by-yigitkonur) using thenpx skillsutility (README.md). - [COMMAND_EXECUTION]: The skill uses shell commands, such as
curl, to perform pre-audit environment checks and verify the availability of target development servers (SKILL.md, Phase 1).
Audit Metadata