skills/polip/opencode-skills/php-pro/Gen Agent Trust Hub

php-pro

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to execute local binaries including vendor/bin/phpstan, vendor/bin/phpunit, and vendor/bin/pest. This functionality involves running code from the local project environment as part of the standard verification workflow.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it processes untrusted user-supplied source code and project configurations. A malicious project could contain instructions or code designed to exploit the agent's command execution capabilities.
  • Ingestion points: The agent ingests PHP source code and project metadata from the local filesystem during architectural analysis and code implementation steps defined in SKILL.md.
  • Boundary markers: Absent. The instructions do not provide explicit delimiters or "ignore embedded instructions" warnings for the agent when reading project content.
  • Capability inventory: The agent is authorized to execute shell commands for testing and static analysis using binaries located in the vendor/bin/ directory (SKILL.md).
  • Sanitization: Absent. There are no specified checks to verify the integrity of the local binaries or sanitize the project content before executing analysis and test commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 07:57 AM
Security Audit — agent-trust-hub — php-pro