cmd-makefile-improve

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard Git commands (git status, git diff, git log, git show) to gather repository context and worktree status. These are routine development operations used to inform the AI's recommendations.
  • [COMMAND_EXECUTION]: The skill executes make help and make -n (dry runs) to inspect the existing Makefile logic and verify that targets function as documented without performing unintended side effects.
  • [DATA_EXPOSURE]: The skill reads local repository files (e.g., AGENTS.md, README.md, Makefile, SKILL.md) and configuration files to understand the project structure. All collected information is presented to the user as part of a proposal and does not involve external network exfiltration or access to sensitive global paths like ~/.ssh or ~/.aws.
  • [PROMPT_INJECTION]: The skill includes strong internal controls and a strict 'Operating contract' that prevents it from modifying files, committing, or installing dependencies without explicit user confirmation (e.g., 'Wait for explicit approval'). It explicitly instructs the agent to 'infer nothing from filenames alone' and 'report evidence... before editing'.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 10:07 AM
Security Audit — agent-trust-hub — cmd-makefile-improve