fix-accessibility-violations

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Chromatic accessibility reports and uses it to directly modify source code. This creates an attack surface where malformed or malicious reports could influence the agent to introduce security vulnerabilities or unintended logic into the codebase.
  • Ingestion points: External build data retrieved via chromatic_list_accessibility_changes and chromatic_get_accessibility_diff tools in SKILL.md.
  • Boundary markers: Absent; the instructions do not include delimiters or warnings to treat external report content as untrusted data.
  • Capability inventory: The skill allows the agent to modify component source files (Step 4) and execute git commit and git push (Step 5).
  • Sanitization: No sanitization or validation of the violation rules or axe documentation content is mentioned.
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform autonomous git operations, including pushing code changes to remote repositories without user confirmation on feature branches.
  • Evidence: Step 5 of SKILL.md explicitly instructs the agent to "commit the fix and push it without asking" on feature branches with open pull requests.
  • Context: While these actions are intended to facilitate the fix-and-verify workflow, high-autonomy write and push operations increase the potential impact of errors or malicious manipulation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 08:30 PM