Scout
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| SCOUT_PROXY | No | Proxy configuration, e.g., http://127.0.0.1:7897 | |
| SCOUT_REDIS_URL | No | Redis URL if using redis cache | |
| SCOUT_SAFESEARCH | No | Safe search level: off, moderate, strict | moderate |
| SCOUT_MAX_RESULTS | No | Maximum number of results | 10 |
| SCOUT_CACHE_BACKEND | No | Cache backend: sqlite or redis | sqlite |
| SCOUT_REQUEST_TIMEOUT | No | Request timeout in seconds | 15 |
| SCOUT_CACHE_TTL_SECONDS | No | Cache TTL in seconds (default 7 days) | 604800 |
| SCOUT_RATE_LIMIT_PER_MINUTE | No | Rate limit per minute | 60 |
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": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| searchC | Run a multi-engine web search and return a ranked, deduplicated link list. Args: translate_to: 翻译目标语言（例如 "zh", "en", "ja"）。需要启用 LLM。 |
| fetchA | Fetch one URL and return reader-mode Markdown of the main content. Args: url: Absolute http(s) URL. extract_code: 提取代码块（默认 False）。 extract_tables: 提取表格数据（默认 False）。 format: "markdown" or "json". |
| fetch_batchA | Fetch multiple URLs in parallel and return their contents. |
| enginesA | List engine names accepted by the |
| cache_searchB | Full-text search across previously fetched and cached pages. |
| semantic_searchB | Search through semantically indexed pages using meaning-based matching. |
| semantic_index_pageA | Fetch a page and add it to the semantic search index. |
| image_searchC | Search for images using SearXNG meta-search engine. |
| researchC | One-shot research: search the web, fetch the top results, return both. |
| summarizeC | Search, fetch, extract key points, and assess credibility. |
| extract_structuredA | Pull JSON-LD, OpenGraph, Twitter cards from a web page. |
| read_docC | Read an http(s) document (PDF) into Markdown. |
| video_searchA | Search for videos on YouTube and Bilibili. Args: query: Search query. platforms: Video platforms to search. Default: ["youtube", "bilibili"]. max_results: Max results per platform. freshness: Time filter. format: "markdown" or "json". |
| video_platformsA | List available video search platforms. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| research_prompt_mcp | Instruct the model to do a thorough, cited research pass. |
| factcheck_prompt_mcp | Instruct the model to fact-check a specific claim. |
| news_brief_mcp | Instruct the model to produce a fresh news brief. |
| compare_sources_mcp | Instruct the model to compare and cross-reference multiple sources. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| cache_stats | Current cache entry counts and configuration. |
| search_history | Recent search queries and their results. |
| engines_list | List of all registered search engines with their capabilities. |
| engines_health | Current health status of all search engines (circuit breaker, rate limit). |
TDQS
Scored across 14 tools
Tools are mostly distinct: search vs. research overlap slightly (research does search + fetch top results), and fetch vs. read_doc both fetch but read_doc is PDF-specific. Most others clearly differentiate.
Naming is inconsistent with mix of verb-first (fetch, search, summarize), noun-first (cache_search, video_search), and plain nouns (engines, research). No uniform pattern but still readable.
14 tools cover a well-scoped domain of web research, fetching, searching, caching, semantic indexing, and summarization. Each tool has a clear purpose and justifies its existence.
Covers core research workflows: search (web, image, video), fetch, read PDF, extract metadata, cache, semantic index/search, summarize. Minor gaps like cache deletion or management, but not essential for primary use.