Security risk report for @forgecat/openai_skills_render-deploy v0.0.8
Source Integrity
Low
Profile is sourced from OpenAI's official skills repository (https://github.com/openai/skills) with clear attribution and Apache-2.0 license.
Metadata includes original commit hash and timestamp, enabling verification of authenticity and integrity.
No evidence of tampering, supply-chain injection, or malicious modification in the declared source.
Agent Intent
Low
Content describes legitimate Render deployment workflows (Blueprint method, Direct Creation via MCP) without instructing the agent to ignore instructions, exfiltrate data, or perform unauthorized actions.
Documentation provides educational guidance on codebase analysis, service configuration, and troubleshooting—all aligned with the stated purpose of deploying applications to Render.
No evidence of prompt injection, system prompt leakage, or guidance poisoning; references to error patterns and debugging are descriptive, not prescriptive for malicious behavior.
Details
Evidence
README and SKILL.md clearly state the skill's purpose: 'Deploy applications to Render by analyzing codebases, generating render.yaml Blueprints, and providing Dashboard deeplinks.'
Prerequisites section instructs users to verify Git remotes and MCP configuration—standard deployment checks, not manipulation.
Error patterns and troubleshooting guides document common failures (port binding, missing env vars, build timeouts) to help users fix legitimate issues.
Permissions
Low
MCP server definition (render: https://mcp.render.com/mcp) is scoped to Render's official API endpoint and requires explicit API key authentication.
No shell, file_write, file_delete, or other high-risk tool categories are declared; the skill operates through Render's MCP interface only.
Authority is narrowly aligned with the stated function: creating and managing Render services, databases, and deployments.
Details
Evidence
MCP configuration specifies a single, known-safe endpoint (https://mcp.render.com/mcp) with Bearer token authentication.
Tool descriptions (list_services, create_web_service, create_postgres, etc.) are limited to Render resource management.
No alwaysApply rules, glob patterns, or unrestricted filesystem/network access.
MCP Risk
MCP-000Medium
The MCP server points to https://mcp.render.com/mcp, which is Render's official hosted MCP endpoint — a known, legitimate service rather than an unknown or attacker-controlled binary.
The server requires a RENDER_API_KEY credential, meaning it has access to the user's full Render account (services, databases, environment variables, logs), which is a meaningful privilege scope.
No hidden instructions are embedded in tool descriptions, and no arbitrary binary execution or unrestricted filesystem access is involved; risk is limited to the external API access scope.
Details
Evidence
MCP config: '"url": "https://mcp.render.com/mcp"' — official Render MCP endpoint
ForgeCat