{"id":45399303,"url":"https://github.com/bit-web24/autotest","last_synced_at":"2026-02-21T19:47:18.971Z","repository":{"id":318314516,"uuid":"1067395249","full_name":"bit-web24/autotest","owner":"bit-web24","description":"An Agentic AI for testing random code","archived":false,"fork":false,"pushed_at":"2026-02-11T17:07:17.000Z","size":380,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"master","last_synced_at":"2026-02-12T01:40:49.757Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/bit-web24.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-09-30T19:50:41.000Z","updated_at":"2025-12-01T19:01:56.000Z","dependencies_parsed_at":"2025-10-06T14:16:26.786Z","dependency_job_id":"6f392b32-8e90-42e3-ae4d-f950385c7bc5","html_url":"https://github.com/bit-web24/autotest","commit_stats":null,"previous_names":["bit-web24/autotest"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/bit-web24/autotest","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bit-web24%2Fautotest","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bit-web24%2Fautotest/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bit-web24%2Fautotest/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bit-web24%2Fautotest/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/bit-web24","download_url":"https://codeload.github.com/bit-web24/autotest/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bit-web24%2Fautotest/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29691081,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-21T18:18:25.093Z","status":"ssl_error","status_checked_at":"2026-02-21T18:18:22.435Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-02-21T19:47:18.861Z","updated_at":"2026-02-21T19:47:18.962Z","avatar_url":"https://github.com/bit-web24.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Autotest\n\nAn Agentic AI for testing random code and write code agentically.\n\n## Architecture \u0026 Design\n\nThis project implements a sophisticated **Supervisor-Worker** agentic workflow using **LangGraph** and **Groq**. The system is designed to autonomously analyze, plan and code software tasks.\n\n### Core Components\n\nThe architecture consists of a central **Supervisor** that orchestrates specialized sub-agents:\n\n1.  **Supervisor Agent** (`agent.supervisor`)\n    *   **Role**: The project manager and orchestrator.\n    *   **Responsibility**: Decomposes complex user requests into sub-tasks, delegates work to the correct sub-agent (`planner` or `coder`), monitors progress, and handles state handoffs.\n    *   **Implementation**: Built using `langgraph_supervisor`.\n    *   **Model**: Defaults to `llama-3.1-8b-instant` via Groq.\n    *   **Memory**: Uses `AsyncSqliteSaver` (SQLite) for persistent state management and checkpointing.\n\n2.  **Planner Agent** (`agent.planner`)\n    *   **Role**: The strategic thinker.\n    *   **Responsibility**: Creates detailed implementation plans, breaks down tasks, and manages the \"todo\" list.\n    *   **Tooling**: Access to the filesystem via MCP (Model Context Protocol).\n    *   **Type**: ReAct Agent.\n\n3.  **Coder Agent** (`agent.coder`)\n    *   **Role**: The executor and engineer.\n    *   **Responsibility**: Writes code, executes commands, fixes errors, and implements the plans generated by the Planner.\n    *   **Tooling**:\n        *   **Filesystem MCP**: For reading/writing files and running commands (`npx @modelcontextprotocol/server-filesystem`).\n        *   **Context7 MCP**: For query based context retrieval (`npx @upstash/context7-mcp`).\n    *   **Type**: ReAct Agent with a specialized system prompt for code analysis and enhancement.\n\n### Data Flow\n\n```mermaid\ngraph TD\n    User([User Request]) --\u003e Supervisor\n    \n    subgraph \"Core Orchestration (LangGraph)\"\n        Supervisor[Supervisor Agent]\n        State[(Shared Graph State)]\n        Persistence[(SQLite Persistence)]\n        Summarizer[Summarizer / Reflector]\n        \n        Supervisor \u003c--\u003e State\n        State \u003c--\u003e Persistence\n        State \u003c--\u003e Summarizer\n    end\n\n    subgraph \"Worker Agents (ReAct)\"\n        Planner[Planner Agent]\n        Coder[Coder Agent]\n    end\n\n    subgraph \"MCP Infrastructure\"\n        FS_MCP[Filesystem MCP Server]\n        Context_MCP[Context7 MCP Server]\n    end\n\n    Supervisor -- \"Delegates Task\" --\u003e Planner\n    Supervisor -- \"Delegates Task\" --\u003e Coder\n    \n    Planner -- \"Read/Write Files\" --\u003e FS_MCP\n    Coder -- \"Read/Write/Execute\" --\u003e FS_MCP\n    Coder -- \"Semantic Search\" --\u003e Context_MCP\n    \n    Planner -- \"Status/Plan\" --\u003e Supervisor\n    Coder -- \"Results/Code\" --\u003e Supervisor\n    \n    Supervisor --\u003e User\n```\n\n\n### Key Features\n\n*   **Model Context Protocol (MCP)**: The agents use MCP to standardize tool access (e.g., filesystem, external context).\n*   **State Management**: Uses `SummaryState` to summarize past interactions and `AsyncSqliteSaver` to persist multiple global states for different chat sessions via unique thread IDs.\n*   **Memory Handoffs**: Sub-agents manage task-specific memory by reading from and writing to the shared graph state, ensuring the Supervisor can transition between planning and coding without loss of context.\n*   **Reflection**: The agents (especially Coder) are designed with a \"reflection\" loop to analyze their own work and correct errors before finishing.\n*   **Pre-Model Hooks**: A `summarization_node` runs before model calls to optimize context window usage.\n\n### Directory Structure\n\n*   `agent/src/agent/supervisor`: Main entry point and orchestration logic.\n*   `agent/src/agent/planner`: Planner agent definition and tools.\n*   `agent/src/agent/coder`: Coder agent definition and tools.\n*   `agent/src/agent/tools`: Shared tool definitions (legacy/utility).\n*   `agent/src/agent/models.py`: LLM configurations (Groq, OpenAI, etc.).\n*   `agent/src/agent/settings.py`: Configuration and environment variables.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbit-web24%2Fautotest","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbit-web24%2Fautotest","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbit-web24%2Fautotest/lists"}