scaffold-project

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell commands to inspect the environment (ls, git), initialize git repositories (git init), and run project validation tools such as tests, linters, and type-checkers as described in Step 3.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted files (e.g., manifests like package.json, Cargo.toml) in the target directory and executes commands derived from them during verification.\n
  • Ingestion points: Target directory inspection commands (ls, git ls-files) and manifest detection logic in SKILL.md.\n
  • Boundary markers: Absent; instructions do not explicitly tell the agent to ignore instructions embedded in the discovered files or manifests.\n
  • Capability inventory: Significant file system write capabilities and shell command execution for project setup and verification across SKILL.md Steps 2 and 3.\n
  • Sanitization: No explicit validation or sanitization of command strings extracted from project manifests before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 02:10 PM
Security Audit — agent-trust-hub — scaffold-project