fcode-cli

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install the @factorialco/fcode-cli package globally using pnpm. This is the official CLI tool for the Factorial Code platform and matches the author's vendor namespace.
  • [COMMAND_EXECUTION]: The documentation describes various commands such as fcode run, fcode push, and fcode http which execute local code or start local servers. These capabilities are the primary intended function of a development CLI and are documented with appropriate usage context.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes tools that ingest untrusted data, such as fcode http for local webhook testing and fcode run with external parameters.
  • Ingestion points: Webhook payloads processed by fcode http and parameter files/JSON strings passed to fcode run (SKILL.md).
  • Boundary markers: The platform uses structured metadata (metadata.json) to define trigger boundaries, but dynamic parameter interpolation is a standard feature.
  • Capability inventory: The tool can execute shell commands via fcode run, perform network sync via fcode push/pull, and start local HTTP servers.
  • Sanitization: The skill advises using webhook.authMode and environment-based secret management to protect local and remote endpoints.
  • [CREDENTIALS_UNSAFE]: The skill provides guidance on managing sensitive data. It explicitly recommends using variables.local.env for local-only secrets to prevent them from being pushed to the cloud or committed to version control. It also notes that sensitive values are replaced with ******** placeholders during synchronization, which is a security best practice for secret redaction.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 12:39 PM