qkb
by miguelarios
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| QKB_CONFIG | No | Path to an alternative config.toml file (points qkb at a different config file entirely). | |
| QKB_RERANK | No | Enable reranking by default (config.toml key: rerank.enabled). | |
| QKB_DB_PATH | No | Path to the SQLite database file (config.toml key: database.path). Default: ~/.local/share/qkb/qkb.db. | |
| QKB_MCP_HOST | No | Host to bind the HTTP MCP server to (config.toml key: mcp.host). Default: 127.0.0.1. | |
| QKB_MCP_PORT | No | Port for the HTTP MCP server (config.toml key: mcp.port). Default: 8181. | |
| QKB_EXPANSION | No | Enable query expansion by default (config.toml key: expansion.enabled). | |
| QKB_VAULT_NAME | No | Vault name used to build obsidian:// links (config.toml key: vault.name). | |
| QKB_VAULT_PATH | No | Path to your Obsidian vault (config.toml key: vault.path). Names exactly one vault; when set, replaces a configured [[vaults]] list. | |
| QKB_OLLAMA_HOST | No | Ollama HTTP API host (config.toml key: embedding.ollama_host). Default: http://localhost:11434. | |
| QKB_EMBEDDING_DIM | No | Embedding dimension (config.toml key: embedding.dimension). Default: 768. | |
| QKB_OPENAI_API_KEY | No | OpenAI API key (environment variable only — never stored in config.toml). Required when embedding.provider is openai. | |
| QKB_WATCH_INTERVAL | No | Seconds between re-index runs in watch mode (config.toml key: watch.interval). Default: 300. | |
| QKB_ALLOWED_ORIGINS | No | Comma-separated browser origins allowed besides loopback; ["*"] disables the check (config.toml key: mcp.allowed_origins). | |
| QKB_EMBEDDING_MODEL | No | Embedding model name (config.toml key: embedding.model). Default: embeddinggemma-300M-Q8_0. | |
| QKB_LOCAL_GGUF_FILE | No | GGUF file name within the repo (config.toml key: embedding.local_gguf_file). Default: embeddinggemma-300M-Q8_0.gguf. | |
| QKB_LOCAL_GGUF_REPO | No | Hugging Face repo for the local GGUF model (config.toml key: embedding.local_gguf_repo). Default: ggml-org/embeddinggemma-300M-GGUF. | |
| QKB_MODEL_CACHE_DIR | No | Directory where downloaded models are cached (config.toml key: embedding.model_cache_dir). Default: ~/.cache/qkb/models. | |
| QKB_OPENAI_BASE_URL | No | Base URL for an OpenAI-compatible /v1/embeddings endpoint (config.toml key: embedding.openai_base_url). | |
| QKB_RERANK_PROVIDER | No | Reranker provider: llama | fake (config.toml key: rerank.provider). | |
| QKB_EMBEDDING_PROVIDER | No | Embedding provider: llama | ollama | openai | fake (config.toml key: embedding.provider). Default: llama. | |
| QKB_EXPANSION_PROVIDER | No | Query expansion provider: llama | fake (config.toml key: expansion.provider). | |
| QKB_EMBEDDING_DOC_TEMPLATE | No | Explicit "{t}"-placeholder document prompt template, overriding the per-model default asymmetric prefixing (config.toml key: embedding.doc_template). | |
| QKB_EMBEDDING_QUERY_TEMPLATE | No | Explicit "{t}"-placeholder query prompt template, overriding the per-model default asymmetric prefixing (config.toml key: embedding.query_template). |
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
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessUnresponsive