forge-new

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to execute a local Node.js script: node ../../../.fullstack-forge/runtime/cli/src/composition-entry.js. This enables command execution using scripts provided by the Fullstack Forge framework.\n- [REMOTE_CODE_EXECUTION]: The skill instructs the agent to load and strictly follow instructions from an external file: ../../../.fullstack-forge/skills/forge-new/SKILL.md. This pattern redirects the agent to execute logic defined outside the skill's own package boundaries.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests logic and configuration from external files, creating a potential surface for indirect prompt injection.\n
  • Ingestion points: ../../../.fullstack-forge/skills/forge-new/SKILL.md, .forge/composition.json.\n
  • Boundary markers: Absent; the agent is instructed to follow the external playbook 'exactly'.\n
  • Capability inventory: Includes subprocess calls (via node) and file system reads.\n
  • Sanitization: There are no specified mechanisms to sanitize or validate the content of the external configuration or playbook files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 06:48 PM
Security Audit — agent-trust-hub — forge-new