fcode-code-validation
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 the agent to install the
@factorialco/fcode-clipackage globally usingpnpm. This is a vendor-owned resource matching the skill's author ('factorialco') and is required for the skill's operation. - [COMMAND_EXECUTION]: The skill utilizes several CLI commands (e.g.,
fcode clone,fcode pull,fcode status) to interact with the Factorial Code platform. These executions are constrained to the intended purpose of auditing and validating application code. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted code and metadata from external workspaces, which presents a surface for indirect prompt injection if the audited content contains malicious instructions designed to subvert the agent's logic.
- Ingestion points: Code files (e.g.,
index.js,main.py), metadata (e.g.,metadata.json,settings.json), and documentation (e.g.,README.md) within the cloned workspace are read during the inventory and catalog check phases. - Boundary markers: No explicit boundary markers or 'ignore' instructions are provided for the ingestion of the external code files.
- Capability inventory: The skill has the ability to read and write files, execute CLI commands, and install packages.
- Sanitization: None detected; the skill performs static checks based on a predefined catalog but does not sanitize the input strings before processing them through the LLM.
Audit Metadata