forge-tenancy

Warn

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a Node.js script located at the relative path ../../../.fullstack-forge/runtime/cli/src/composition-entry.js with parameters that include repository roots and JSON flags.
  • [REMOTE_CODE_EXECUTION]: The instructions mandate that the agent read and follow a 'canonical playbook' from a relative path (../../../.fullstack-forge/skills/forge-tenancy/SKILL.md), effectively delegating the agent's logic to external, non-static files.
  • [PROMPT_INJECTION]: The skill has an indirect injection surface where it ingests data from .forge/composition.json to determine which instruction paths to load into the agent context. Ingestion points: .forge/composition.json. Boundary markers: Absent. Capability inventory: Execution of Node.js scripts and file system reading across relative directory structures. Sanitization: None observed in the provided instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 6, 2026, 06:48 PM
Security Audit — agent-trust-hub — forge-tenancy