sruja
Active·★ 24·Apache-2.0·Updated 2026-10-19
Sruja is an autonomous, CLI-first coding agent designed to scan repositories, generate code edits, and rigorously verify them. It employs an independent deterministic grader to ensure code quality and alignment with intent, preventing self-grading by the AI actor.
#Autonomous Agent#Code Generation#Code Verification#Developer Tools#CLI#LLM Integration#Code Refactoring#Software Engineering
01
Features
01Autonomous Agent Loop: Operates a closed-loop system for observation, action, verification, critique, and replanning.
02Independent Deterministic Grader: Utilizes tools like `drift`, `lint`, `verify-task`, and `intent check` to independently verify code edits.
03Editor Integration via MCP: Provides grounded context and post-edit verification gates for various AI editors (e.g., Cursor, Copilot).
04Knowledge Capture and Retrieval: Allows ingesting design documents and decisions, and retrieving task-specific context with blast radius and dependencies.
02
Compatibility
CLI
Interface
Verified via docs
Linux / MacOS
Operating System
Verified via docs
AI Editors (Cursor, Copilot, Claude, Windsurf)
Integration
Verified via docs
03
Quick start
1
$ curl -fsSL https://sruja.ai/install.sh | bash
04
Use cases
↳Automated Code Refactoring: Autonomously refactor code (e.g., auth boundaries) while ensuring tests and structural integrity are maintained.
↳Post-Edit Verification: Verify code alignment after manual or AI-driven edits using deterministic checks for structure and intent.
↳Contextual Briefing for LLMs: Provide file-scoped briefings with dependencies and relevant decisions to host LLMs or agents before making changes.
↳Enforcing Design Intent in CI: Use `repo.sruja` with `lint` and `drift` in CI to enforce architectural boundaries and design decisions.
05
Alternatives
pi-code★ 36
Claude Code experience for the pi coding agent: reads your .claude config (rules, commands, skills, hooks, output styles, MCP, agents) and adds todo, checkpoints, memory, web search, and subagents.
refact★ 3.5k
AI Agent that handles engineering tasks end-to-end: integrates with developers’ tools, plans, executes, and iterates until it achieves a successful result.
Benzi★ 169
Benzi is AI-native code intelligence infrastructure. Claude Code greps; Cursor embeds; Aider maps signatures; Benzi resolves — and answers in O(1). Every language runs its own tree-sitter grammar into the same query map.
RoslynMcpServer★ 31
Model Context Protocol server for Roslyn-powered C# refactoring operations
ios-agent-skill★ 40
Swift source, Apple guides and MCP tools for AI coding agents. Review iOS code, scaffold apps, and build/run/preview with Xcode Simulator.
ThreeBrowserStudio★ 24
LLM-first Three.js WebGPU scene and animation authoring with a persistent native viewport and typed MCP editing
Comments
Log in to leave a comment
No comments yet. Be the first!