Report Abuse

Basic Information

PeterCat is a solution for creating, deploying and running repository-specific conversational Q&A agents aimed at community maintainers and developers. The project provides an agent configuration system and an application SDK so teams can generate an intelligent support bot for a given GitHub repository by supplying the repo address or name. It automates generation of bot metadata such as prompts, names, avatars, opening lines and tool sets, and orchestrates ingestion of repository content and resolved issues into a knowledge base. The repo includes self-hosted deployment options with guides for local and AWS+Supabase setups, environment variable conventions for LLMs and services, and packaging targets such as npm, Docker and PyPI for distribution.

Links

Categorization

App Details

Features
Preconfigured agent workflow that can auto-create a repository Q&A bot from a GitHub repo. Automated knowledge ingestion pipeline that loads repository Markdown and issues then processes them via load -> split -> embed -> store to build a retrievable knowledge base. Multi-platform integration options including a front-end conversation SDK and a GitHub App for one-click installation. Self-hosting guides and architecture for local and AWS+Supabase deployments. Support for multiple LLM and RAG service integrations through environment variables including OpenAI, Gemini, Tavily and a WHISKER RAG endpoint. Developer tooling and scripts using yarn for bootstrapping, running client/assistant/server, and building Docker and packaging artifacts.
Use Cases
PeterCat helps open source projects and teams provide faster, consistent technical support by converting repository content and community discussions into a searchable knowledge base used by a conversational agent. It reduces manual setup by auto-generating bot configuration and orchestrating ingestion tasks, so maintainers can deploy a repo-specific assistant with minimal steps. Multi-platform SDKs and a GitHub App make it straightforward to expose the assistant on project websites or inside GitHub workflows. Self-hosting options and environment-driven integrations let organizations control deployment, choose LLM/RAG providers and meet enterprise needs. Bundled developer scripts and packaging targets simplify testing, local development, Docker builds and distribution.

Please fill the required fields*