Skip to main content
Glama
This connector has been deprecated

Replaced by echopai-mcp.

Duplicate listing; superseded by io.github.echopai/echopai-mcp (official MCP Registry entry).

news

SUPPLEMENTARY news / market briefs (short-horizon event stream) — use only to fill gaps not covered by views, the primary research source. Default fields=compact (title + ai_summary + ~300 chars); full only when needed. Flash news has title: null — read display_title (falls back to the first sentence), or pass has_title=true to drop untitled items. Modes: code (canonical, A-share / HK / US) → news mentioning that security; query (free text, all terms required, newest first); neither → recent feed. A canonical_code passed as query will not match — use code. A+H: pass the A-share canonical.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "news"
  }
}

Input Schema

NameRequiredDescriptionDefault
toNoHistory range end (inclusive, China date; defaults to today). Use with from.
_jqNoJMESPath,对 data 求值(写 items[*].id)
codeNoFilter by security canonical_code (e.g. SSE:600519, HK:00700). Takes precedence over `query`.
fromNoHistory range start (China date YYYY-MM-DD). Range mode: needs code or query; Pro-only up to 365 days back.
hoursNoRolling-feed lookback in hours (default 24 — much narrower than views). For longer history use from/to with code or query. The `limit` cap applies after this window, newest first, so a dense feed can shrink the covered span far below what you asked for — read `actual_window` in the response for the range actually returned, and page with `cursor` (from `next_cursor`) if it falls short.
limitNoMax items per page
queryNoFree-text query (Chinese or English) matched against news content; omit for time-window feed. Terms are AND-ed — a document must contain every term — and results stay in strict reverse-chronological order, so extra terms narrow rather than broaden. Pass 2-3 distinctive terms; for relevance-ranked recall that matches any term, use the `search` tool instead. Chinese needs no spaces, and AND/OR/quotes are treated literally rather than as operators. Do NOT pass canonical_code here — use `code` instead.
_spillNotrue=全量落盘,只回摘要与取数地址
cursorNoOpaque snapshot pagination cursor from the previous response's `next_cursor` (freezes time window & sort anchor; preferred over `offset`).
fieldsNoResponse shape: compact (default — title + ai_summary + first ~300 chars) or full (complete content)
offsetNoLegacy pagination offset (prefer `cursor` from `next_cursor` to avoid duplicates/skips on a live stream).
_fieldsNo逗号分隔的直接字段;有 items 时投影每行,否则投影 data;外层字段或嵌套结构用 _jq
_intentNo用户原问题一句话,仅分析用,可留空
has_titleNotrue = only items with a headline. Flash news has none (often ~40% of the feed), so pass this when filtering or summarising by headline; otherwise read `display_title`, which falls back to the first sentence.
_max_bytesNo响应字节上限,默认约 50 KB
_max_bytes_modeNo正文超限处理,默认 truncate 保留前缀

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • addedInput schema / properties / cursor
      Added value: +{
      +  "description": "Opaque snapshot pagination cursor from the previous response's `next_cursor` (freezes time window & sort anchor; preferred over `offset`).",
      +  "type": "string"
      +}
    • changedInput schema / properties / hours / description
      Previous value: -"Rolling-feed lookback in hours (default 24 — much narrower than views). For longer history use from/to with code or query. The `limit` cap applies after this window, newest first, so a dense feed can shrink the covered span far below what you asked for — read `actual_window` in the response for the range actually returned, and page with `offset` if it falls short."New value: +"Rolling-feed lookback in hours (default 24 — much narrower than views). For longer history use from/to with code or query. The `limit` cap applies after this window, newest first, so a dense feed can shrink the covered span far below what you asked for — read `actual_window` in the response for the range actually returned, and page with `cursor` (from `next_cursor`) if it falls short."
    • changedInput schema / properties / offset / description
      Previous value: -"Pagination offset for paging through long history ranges."New value: +"Legacy pagination offset (prefer `cursor` from `next_cursor` to avoid duplicates/skips on a live stream)."
  2. First observed

TDQS

A4.5/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations exist, so the description carries the full load and does well: it discloses the compact-by-default response shape, that ~40% of flash items have `title: null` and how to handle it (`display_title` fallback or `has_title=true`), and that the `limit` cap is applied after the window so the effective span can be shorter than requested. It does not mention auth tiering beyond the 'Pro-only' history caveat, and output item structure is only partially described, keeping it from a 5.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the supplementary-status caveat, then defaults, then modes, then edge cases — a sensible order for an agent scanning quickly. It is dense but nearly every clause carries operational information; the parenthetical metric re-assertion ('much narrower than views') is mild redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 16-parameter tool with no annotations and no output schema, the description covers the hard parts: mode selection, precedence, default response shape, the null-title trap, window shrinkage, and pagination recovery via `next_cursor`. Remaining gaps are minor — return item fields beyond `display_title`/`actual_window`/`next_cursor` and any tier/rate constraints are only partly implied.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is already 100%, so the baseline is 3; the description earns an increment by explaining precedence and interaction the schema cannot ('code' takes precedence over 'query'; a canonical_code in `query` will not match) and by justifying the `fields` default. It also steers away from `offset` toward `cursor`, which the schema only weakly signals.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb+resource framing: 'SUPPLEMENTARY news / market briefs (short-horizon event stream)' and immediately positions it against `views` as the primary research source. The mode enumeration (`code` → news mentioning a security; `query` → free text; neither → recent feed) makes the retrieval semantics unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit when-to-use ('use only to fill gaps not covered by `views`'), an explicit alternative for a different need ('for relevance-ranked recall that matches any term, use the `search` tool instead'), and explicit exclusions ('A canonical_code passed as `query` will not match — use `code`'). The A+H instruction (pass the A-share canonical) resolves an otherwise ambiguous case.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources