agent-army

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill orchestrates various shell commands to manage work safety and verification. This includes version control operations (git status, checkout, diff), file analysis (wc, grep, glob), and project-specific build commands (e.g., npm run build, Makefile). These operations are consistent with the skill's stated purpose of codebase modification.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes local codebase files as primary input for specialist sub-agents. It lacks explicit sanitization or boundary markers (such as XML tags or explicit 'ignore instructions' warnings) when interpolating file content into agent prompts. This creates a surface where malicious instructions embedded in a project's source code could potentially influence agent behavior, though the skill includes a 'phantom-completion' check using git diff to verify actual work performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 07:25 AM
Security Audit — agent-trust-hub — agent-army