troubleshoot

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell commands to verify plugin integrity and workspace health, including checking for directory existence and validating JSON/YAML structures using Python and Bash scripts.\n- [EXTERNAL_DOWNLOADS]: The documentation references the installation of 'pptxgenjs' via npm to resolve slide generation issues. This is a well-known package and is not automatically installed or executed by the skill.\n- [PROMPT_INJECTION]: The skill reads and processes 'SKILL.md' and '.local.md' files from the local environment to perform diagnostics. This ingestion of untrusted file content creates a surface for indirect prompt injection.\n
  • Ingestion points: Reads content from plugin definition files (SKILL.md) and user state files (.local.md) in the .claude directory.\n
  • Boundary markers: The instructions do not define clear delimiters or use specialized prompts to ignore potential instructions embedded within the files being diagnosed.\n
  • Capability inventory: The skill has access to Bash and file reading tools, which could be misused if the agent obeys instructions found within analyzed files.\n
  • Sanitization: There is no evidence of sanitizing the content of the analyzed files before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 01:51 PM
Security Audit — agent-trust-hub — troubleshoot