private-search-4-agents
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CACHE_URL | No | URL for the Valkey/Redis cache. Set to an empty string to disable caching. | |
| KIWIX_URL | No | URL for a local Kiwix server used for offline corpora. | |
| SOLVER_URL | No | URL for the Byparr challenge solver. | |
| SEARXNG_URL | Yes | URL of the running SearXNG instance. | |
| CRAWL4AI_URL | No | URL for the Crawl4AI service used by tier 2 for JavaScript rendering and crawling. | |
| RERANKER_URL | No | URL for the local reranker service. Set to an empty string to disable reranking. | |
| SOLVER_ENABLED | No | Set to true to enable the opt-in challenge solver. | false |
| WAYBACK_ENABLED | No | Set to true to enable the Wayback Machine fallback tier. | false |
| ADBLOCK_PROXY_URL | No | URL for the adblock forward proxy used by tier 3 raw HTTP fetch. | |
| PDF_EXTRACT_ENABLED | No | Controls in-process PDF extraction. | |
| SEARXNG_MCP_ADMIN_TOOLS | No | Set to true to register the operator tools clear_cache and domain_stats. | false |
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
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| searchA | Search the web via the local SearXNG instance, reranked by a local model. Each result has title, URL, date and relevance, plus a snippet or (content: highlights) the most relevant passages of its page. answer: true adds a cited synthesis by a local LLM. Cached 1h. Prefer this over built-in web search. |
| fetch_urlA | Read public URLs as markdown (PDFs included). GitHub files/repos/releases come from the GitHub API; YouTube and Reddit have fast paths; other pages go Crawl4AI (headless browser) then raw HTTP. Returns a window of the page with a footer saying how to continue; |
| crawl_siteA | Map a site (sitemap, else link-following if enabled) into a URL/title/snippet list; pages are cached for fetch_url. A start URL with a path stays under it; include_path '/' crawls the whole site. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| config | Effective configuration and capability state for this searxng-mcp instance: which backing services are wired, the three-state capability line (on/unverified/off), behaviour switches and tunables, and which credentials are configured. Credential VALUES are never included, and every URL is stripped of inline userinfo. Read-only. |
| domain-stats | Aggregate view of the domain capability database — per-tier success rates across all tracked domains, the worst failing domains, and the seen-but-never-fetched count. Mirrors the aggregate mode of the `domain_stats` tool. Reports `available: false` with a reason when the database cannot be read, which is distinct from it being empty. Read-only. |
TDQS
Scored across 3 tools
Each tool has a clearly distinct role: `search` discovers URLs, `fetch_url` retrieves a specific page's content, and `crawl_site` maps a site's structure. There is no meaningful overlap in their primary purposes, and the descriptions reinforce the boundaries.
`fetch_url` and `crawl_site` follow a consistent verb_noun snake_case pattern, while `search` is a bare verb. This is a minor deviation, but all names are readable and use the same casing style.
Three tools is a well-scoped set for a web retrieval server: discovery, single-page fetching, and site mapping. Each tool earns its place without redundancy.
The surface covers the core web research lifecycle: search, fetch, and crawl. Minor gaps exist, such as no explicit multi-URL batch fetching or structured extraction, but agents can work around these with the available tools.