pre-deploy-checklist

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it performs deep analysis on an untrusted codebase, which can contain malicious instructions intended to manipulate the agent's behavior during Phase 2 or Phase 3.\n
  • Ingestion points: Automated mapping of the project structure in Phase 1 and detailed file exploration by domain subagents in Phase 2 (SKILL.md and references/domain-prompts.md).\n
  • Boundary markers: The skill does not implement delimiters or explicit instructions to ignore embedded commands when processing project files.\n
  • Capability inventory: The skill possesses the Bash (for discovery), Task (for spawning subagents), and Write (for generating the checklist) capabilities.\n
  • Sanitization: No validation or sanitization is performed on the content retrieved from the local project files before it is processed by the AI agents.\n- [COMMAND_EXECUTION]: The skill uses the Bash tool to execute find commands to map the directory structure. While used for discovery, the execution of shell commands on an untrusted directory structure is a notable capability that requires caution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 11:25 AM
Security Audit — agent-trust-hub — pre-deploy-checklist