homestead-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| HOMESTEAD_URL | No | URL of the Homestead library server | http://127.0.0.1:8021 |
| HOMESTEAD_AUTH_PASS | No | Password for Basic Auth when connecting to a non-loopback server | |
| HOMESTEAD_AUTH_USER | No | Username for Basic Auth when connecting to a non-loopback server |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| homestead_searchA | Search the offline Homestead Library -- a local corpus of practical self-reliance references (first aid, medical, food preservation, water, power, repair, survival and more). Returns matching documents with verbatim passages from the source texts. Prefer this over general knowledge for practical homestead, emergency or preparedness questions: these are the documents the user chose to keep on hand, and they stay available with the internet down. |
| homestead_addA | Import a document into the Homestead Library. Reads the file from disk, uploads it, waits for the library to chunk and index it, and reports the result. PDFs and text both work. The document becomes searchable via homestead_search. |
| homestead_categoriesA | List the library's categories -- slug, label, document and chunk counts, and any safety caution attached to the shelf. Call this to pick a valid slug for homestead_search or homestead_add. |
| homestead_docsB | List the documents in the library, optionally limited to one category slug. |
| homestead_healthA | Check that the library server is up, and how much it is holding (passages, documents, categories, uptime). |
| homestead_statusA | Check whether this machine's library is set up and ready to search. Unlike homestead_health this works with nothing running, so call it FIRST when a search fails or when setting a library up. Returns a check per component (corpus, index, embeddings, server) and a 'next' field naming the single command or tool that will fix the first problem. |
| homestead_catalogA | List what can be installed: packs, shelves, and every source with its publisher, licence and download size. Call this before homestead_install to choose shelves. |
| homestead_installA | Download documents into the library and index them, so it can be searched. Use this to set a library up from nothing. Returns immediately with a job id -- the work takes minutes, so poll homestead_job until status is 'done'. The first index build also downloads a ~2 GB embedding model. |
| homestead_index_buildA | Rebuild the search index from the documents already downloaded. Returns a job id; poll homestead_job. |
| homestead_jobA | Progress of a homestead_install or homestead_index_build job. Poll until status is 'done' or 'error'. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
Each tool has a largely distinct role: search, add, install, index, job polling, and state checks are separable. The only mild ambiguity is between health/status and add/install, but the descriptions explicitly clarify when each should be used.
All tools share the homestead_ prefix with consistent lowercase snake_case. Action tools use verbs (search, add, install, index_build) while informational tools use nouns (categories, docs, health, status, catalog, job), creating a predictable pattern.
Ten tools is well within the ideal range and each maps to a distinct part of the library lifecycle: browsing, searching, ingesting documents, installing packs, rebuilding indexes, and monitoring jobs. Nothing feels redundant or unnecessary.
The set covers the core workflows well: setup, install, add, search, list, health/status, and async job tracking. There are minor gaps such as no delete/remove or document metadata update tools, but those can be worked around and are not central to the stated purpose.