no-mistakes

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the no-mistakes CLI family. It interpolates user-controlled text directly into command-line arguments.
  • Evidence: no-mistakes axi run --intent "<what the user set out to accomplish>" and no-mistakes axi respond --action fix --findings <id> --instructions "<optional guidance>".
  • The tool no-mistakes is a vendor-provided CLI for managing code quality pipelines.
  • [PROMPT_INJECTION]: An indirect prompt injection surface exists where untrusted user input is passed to system-level commands.
  • Ingestion points: User-provided task text and intent descriptions provided during invocation or as guidance for fixes (found in SKILL.md).
  • Boundary markers: The instructions recommend wrapping interpolated user strings in double quotes (e.g., "<task>"), which provides a basic delimiter but does not prevent command injection if the input contains nested quotes or shell metacharacters.
  • Capability inventory: The skill has the capability to execute shell commands (no-mistakes), perform git operations (git status, git commit, git push), and interact with remote CI/CD systems via PR creation.
  • Sanitization: There are no instructions for the agent to sanitize, escape, or validate the user-provided text before passing it to the shell execution environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 12:54 AM
Security Audit — agent-trust-hub — no-mistakes