arc42-lint

Warn

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script scripts/arc42-lint.py. This poses a risk if an attacker places a malicious script at that location in a project, as the agent will execute it with the user's permissions.
  • [PROMPT_INJECTION]: The skill processes content from untrusted documentation files, creating a surface for indirect prompt injection.
  • Ingestion points: Markdown files (.md) located at the user-specified path.
  • Boundary markers: No explicit delimiters or warnings to ignore instructions embedded in the documentation are provided.
  • Capability inventory: The agent can execute shell commands (python), read files, and write/modify files (as offered in Step 4).
  • Sanitization: No sanitization or validation of the documentation content is performed before processing or acting upon it.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 10, 2026, 11:22 AM
Security Audit — agent-trust-hub — arc42-lint