Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.unideploy.in/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

When you run a scan, three agents work together to analyse your code and — if you trigger a fix — create a pull request with patches applied.

​
The three agents

​
Orchestrator

The Orchestrator is the root agent. It receives your project manifest (framework, file list, key files), decides which sub-agents to invoke, and assembles the final response. You never interact with it directly.

​
AnalyzerAgent

The AnalyzerAgent runs the 13-rule security checker against your codebase. It produces a structured JSON array of findings — each with a rule ID, severity, file location, description, and auto-fix flag. It also runs the grading algorithm to produce a letter grade (A–F). On paid plans, the AnalyzerAgent also passes each finding through a language model (Gemini 2.0 Flash) to generate a plain-English remediation plan with code examples.

​
FixAgent

The FixAgent is invoked only when you explicitly request fixes (via the dashboard or unideploy fix). It takes the findings produced by the AnalyzerAgent, generates file patches using Gemini 2.5 Pro, applies them to a sandboxed clone of your repository, and creates a GitHub pull request via the Composio GitHub integration. The FixAgent only modifies files and only for findings marked auto_fixable: true.

​
Where agents run

Agents are deployed on Vertex AI Agent Engine (Google Cloud) and are invoked by the backend API. You do not need a Google Cloud account to use UniDeploy — this is infrastructure that UniDeploy manages on your behalf.

​
Sandbox isolation

Code is processed inside an ephemeral sandbox that is destroyed after each session. The FixAgent applies patches to a sandboxed clone, not to your live repository. Changes reach your repo only via a pull request, which you review and merge.