Skip to content
Froggy Softworks — Practical, thoughtful software focused on you.

Froggy Context Hub setup guide

Froggy Context Hub is a local AI gateway that gives your apps access to private knowledge, MCP tools, prompt profiles, and grounded context — without rebuilding RAG for every client. Install from GitHub, configure in the UI, then point your existing OpenAI or Ollama clients at Froggy.

Download

This is the official getting-started guide for Froggy Context Hub. For product overview and support, see the product page or contact FroggySoftworks.

Prerequisites

  • Windows — download the installer from the download page
  • An upstream LLM — OpenAI-compatible API, or a local runtime such as Ollama or LM Studio
  • Documents to index — folders or files for your namespaces (Markdown, PDF, plain text, and more)
  • Optional: MCP clients (Cursor, Claude Desktop) if you want rag_search over your corpora

Installation

Download the latest Windows installer from the download page (currently v1.6.13).

  1. Run the setup installer (.exe or MSI)
  2. Complete the installer — Froggy Context Hub runs locally with auto-updates enabled
  3. Launch the app from the Start menu

Configuration

Most settings are configured in the desktop app under Settings — not via hand-edited config files for day-to-day use.

  • General — project paths and defaults
  • Chunking — how documents are split and embedded
  • Retrieval — top-K, algorithms, always-inject behavior
  • LLM passthrough — inbound listener ports and upstream model URLs
  • MCP Server — inbound rag_search for IDE clients
  • External MCPs — third-party tool servers wired into passthrough chat
  • Web search — optional web results blended with corpus retrieval

Default ports (configurable): Direct RAG REST on 3000, inbound MCP on 3100, plus separate OpenAI- and Ollama-compatible passthrough listeners.

Adding documents

Use the Ingestion tab to point Froggy at folders or files and pull them into a namespace (for example work-docs, personal-notes).

  1. Create or select a namespace
  2. Ingest folders or files — embeddings are built automatically
  3. Confirm indexing finished in the Server / vector store views before querying

Each namespace is an isolated corpus: its own documents, embeddings, and retrieval scope.

Choosing model providers

Froggy sits in front of your LLM and exposes two familiar HTTP APIs. RAG runs the same pipeline on both — only the wire format differs.

  • OpenAI-compatible — point SDKs and clients at Froggy’s OpenAI listener (e.g. /v1/chat/completions)
  • Ollama-compatible — point Ollama-oriented tools at Froggy’s Ollama listener (e.g. /api/chat)

Configure upstream URLs and API keys under Settings → LLM passthrough. For local models, point upstream at your Ollama or LM Studio endpoint.

Running the app

Start Froggy Context Hub from the Start menu. The Server tab shows status for:

  • Direct RAG REST — corpus search, ingest, and admin API
  • Passthrough — OpenAI- and Ollama-compatible inbound listeners
  • MCP Server — rag_search for external MCP clients
  • External MCPs — third-party tools injected into chat

Enable the listeners you need in Settings, then aim your client at Froggy’s host and port using the same paths you would use against vanilla OpenAI or Ollama — except the base URL now points at Froggy.

Using RAG

When a compatible request hits Froggy, it searches the active namespace, retrieves relevant chunks, optionally blends web search or MCP tool results, injects context into the prompt, and forwards to your upstream LLM.

Ways to use it:

  • Passthrough — point existing OpenAI or Ollama clients at Froggy
  • Direct RAG REST — search and ingest over HTTP
  • MCP — call rag_search from Cursor or Claude Desktop
  • In-app tests — run search and passthrough exercises in the Local API tests tab

Always verify important answers against source documents — retrieval quality depends on your corpus, chunking, and prompts.

Troubleshooting

Indexing fails or hangs
Check disk space, file permissions, and embedding provider connectivity. Review in-app logs and the Server tab.
No results or poor answers
Confirm the right namespace is active and documents finished indexing. Tune chunking and retrieval in Settings.
Client cannot reach Froggy
Verify the correct listener is enabled and your client uses Froggy’s port, not the upstream model’s port directly.
Provider / upstream errors
Check upstream URL, API keys, and model names under LLM passthrough settings.
Bugs and feature requests
Contact FroggySoftworks with steps to reproduce, versions, and error messages.

Known limitations

  • Windows installer is the primary distribution today
  • Retrieval quality varies with document quality, chunking, and namespace design
  • Generated answers should be validated for critical use cases
  • Large corpora may require tuning chunk size, hardware, and retrieval settings

For product overview, see the Froggy Context Hub product page.