scythe-context-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GEMINI_API_KEY | Yes | Your Gemini API key (or bearer token for proxy). | |
| GEMINI_BASE_URL | No | Base URL for Gemini-compatible proxy. | |
| GEMINI_AUTH_MODE | No | Authentication mode: x-goog-api-key, bearer, or query. | |
| GEMINI_OUTPUT_DIMENSIONALITY | No | Output dimensionality for embeddings. | 1536 |
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 | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| repo_doctorB | Run local diagnostics for runtime, native modules, config, WSL interop, and index health without calling embedding APIs. |
| repo_index_statusC | Show Scythe Context configuration and current index status. |
| gemini_embedding_probeB | Send one embedding request and return diagnostics for official Gemini or proxy compatibility. |
| repo_reindexC | Scan a project, write metadata, and optionally index embeddings. |
| repo_semantic_searchA | Debug raw semantic or hybrid ranking over indexed chunks. Prefer repo_context_pack for normal lookup; use paths_only to scout cheaply and snippets only when scores or fuller snippets are needed. |
| repo_related_filesC | Show compact symbols, imports, and reverse imports for one indexed file after a candidate path is known. |
| repo_context_packB | Preferred task-oriented code lookup for unknown file locations. Defaults to paths_only; use compact for short snippets and snippets for ranking/debug detail. Set mode=keyword for local lookup without embedding requests. |
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 7 tools
Most tools have clearly distinct purposes, with explicit guidance that repo_context_pack is preferred for normal lookup and repo_semantic_search is for debugging rankings. The main possible confusion is between the two lookup tools, and secondarily between repo_doctor and gemini_embedding_probe, but descriptions mitigate this well.
All names use snake_case, and six of seven share a repo_ prefix, creating a mostly predictable pattern. The exception is gemini_embedding_probe, and the names mix noun phrases with action-oriented labels, preventing a perfect score.
With 7 tools, the set is well-scoped for a context/indexing MCP server. Each tool has a clear role spanning diagnostics, indexing, status, search, and related-file lookup without obvious redundancy.
The server covers diagnostics, indexing lifecycle, status, semantic search, context packing, and related-file inspection, which addresses the core workflows. Minor gaps exist, such as no explicit index deletion/reset or project listing tool, but agents can likely work around them.