code-simplifier

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content by reading local code files and project instructions (AGENTS.md) which could contain malicious prompts.
  • Ingestion points: The agent reads recently modified files and AGENTS.md as specified in the workflow steps of SKILL.md.
  • Boundary markers: No specific delimiters or instructions are provided to the agent to distinguish between its own instructions and the content of the files it processes.
  • Capability inventory: The skill can modify files and execute verification commands such as tests and linting (SKILL.md).
  • Sanitization: No sanitization or validation of the ingested code content is defined.
  • [COMMAND_EXECUTION]: The skill is instructed to run local commands for scope definition and code verification.
  • Evidence: SKILL.md directs the agent to use git diff and git status to identify changes, and to execute targeted tests, linting, and build checks during the verification phase.
  • [EXTERNAL_DOWNLOADS]: The skill contains a reference to an official GitHub repository.
  • Evidence: references/upstream-agent.md links to the official Anthropic Claude plugins repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 04:21 PM
Security Audit — agent-trust-hub — code-simplifier