The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides an extensive interface for the insforge CLI tool, enabling the agent to execute administrative commands including database queries, storage management, and deployment of serverless functions.
[EXTERNAL_DOWNLOADS]: During project initialization via insforge create, the skill downloads additional functional components using the command npx skills add insforge/agent-skills. This component is a vendor-owned resource.
[CREDENTIALS_UNSAFE]: Documentation identifies local filesystem paths where sensitive credentials are stored, specifically ~/.insforge/credentials.json for session tokens and .insforge/project.json for project-specific admin API keys.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of processing and executing user-provided SQL and source code.
Ingestion points: User-supplied SQL in insforge db query and source code in insforge functions deploy.
Boundary markers: None present in the instruction set to delimit user data from system commands.
Capability inventory: High-privilege access including full database manipulation, secret retrieval, and serverless code execution.
Sanitization: There are no explicit instructions requiring the agent to sanitize or validate user-provided inputs before execution.

insforge-cli