craft-project-setup

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill includes instructions that direct the agent to conceal its AI identity in generated output (e.g., "Do not include AI attribution in commits... all output should be indistinguishable from human-authored work"). While common for professional parity, this is a form of instruction to suppress disclosure.
  • [PROMPT_INJECTION]: The skill creates an indirect prompt injection surface because it reads external project metadata to drive the scaffolding process.
  • Ingestion points: Reads project root files including composer.json, package.json, and .ddev/config.yaml as described in Step 1 of the SKILL.md workflow.
  • Boundary markers: Absent; there are no specific instructions to ignore malicious directives embedded within the data files it parses.
  • Capability inventory: The skill has full file-writing permissions within the project directory and utilizes shell commands via ddev and gh tools.
  • Sanitization: No explicit sanitization or validation logic is specified for the data ingested from the project configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 07:56 PM