forma-cli

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs the agent to documentation for installing software from the vendor's official GitHub repository at choral-io/choral-forma.
  • [COMMAND_EXECUTION]: The skill provides instructions for executing various CLI commands through the forma binary to inspect workspace health, view configurations, and manage skills.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external workspace files and schema definitions, creating a potential surface for instructions embedded in untrusted data to influence agent behavior.
  • Ingestion points: Reads workspace files via forma workspace explain and retrieves documentation via forma docs get.
  • Boundary markers: Explicitly instructs the agent to stay within the user's approved write scope and requires human approval for modifications.
  • Capability inventory: Provides tools for reading workspace health, inspecting configuration, and creating files (via forma create).
  • Sanitization: No specific programmatic sanitization of workspace content is described beyond the use of structured JSON output for tool results.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 02:37 PM
Security Audit — agent-trust-hub — forma-cli