Sliq

Google news search

google_news_search

Read-only

A find-companies News-mode discovery source, used alongside search_news_events, tavily_search, and exa_search_news — the News-mode skill owns the routing (which signals fan out here). Its edge is Google's local and trade-press index: hyper-local expansion stories, plant incidents within hours of publication, and small-market executive moves the other sources' indexes miss. Returns raw articles, not company rows — extract the company from each article's title/snippet the same way as tavily_search results.

Compose KEYWORD-style queries — the terms a journalist's headline would contain, e.g. 'appoints new chief marketing officer' or 'breaks ground manufacturing facility'. Google matches keywords, not meaning: full-sentence semantic phrasing ("A company that recently...") returns ZERO results. Geographic scoping via keywords is brittle (a location term matches any article containing the word, wherever the company is) — prefer scoping by event keywords and filtering location downstream. One query per signal, all in one call (the tool fans them out in parallel; one tool call is one LLM round).

Expect the same story to arrive several times from different outlets — Google clusters dilute a page's effective yield to ~5–7 distinct companies. Dedup is on exact URL here; cross-outlet same-story dedup happens downstream at record_search_results.

Articles are recency-bounded server-side by Google's coarse day/week/month/year buckets, using the smallest bucket covering lookback_days (default 30) — a bucket can exceed the window, so also check each row's published_at when the window matters. A news scan passes the window from its trigger prompt.

Costs 0.05 Sliq credits per query that returns without error; a failed query is free. Companies you record from its articles are not charged. Soft-fails on insufficient credits: the charge is capped at the remaining balance. Dict with articles array and count. Each article row is {title, url, source, snippet, date, published_at, query} — source is the publisher name, date is Google's relative publish string ("2 days ago"), published_at is a coarse UTC ISO estimate derived from it (None when unparseable), and query names the originating query. Rows are deduped by URL across queries.

If every query failed with an upstream error and nothing landed, also google_news_available: False plus a human-readable error — a source outage, NOT a genuinely empty feed. Partial failure (some queries errored, others returned) stays silent and reports only the articles.

On insufficient Sliq credits the call is refused before running — it raises InsufficientCreditsError instead.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "queries": "<queries>"
    },
    "name": "google_news_search"
  }
}

Input Schema

Name	Required	Description
`queries`	Yes	One keyword-style news query per signal. Pass a list of one for a single-signal lookup.
`on_behalf_of`	No	A teammate's email, or an agency client's from list_agency_clients, to run this call in that account: their agents, settings, connected accounts, and credits. Omit it to act on your own account.
`lookback_days`	No

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedNov 6, 2026

changedInput schema / properties / on_behalf_of / description

Previous value: -"An agency client's email, from list_agency_clients, to run this call in that client's account: their agents, settings, connected accounts, and credits. Omit it to act on your own account."New value: +"A teammate's email, or an agency client's from list_agency_clients, to run this call in that account: their agents, settings, connected accounts, and credits. Omit it to act on your own account."

Changed1 schema field changedOct 19, 2026

addedInput schema / properties / on_behalf_of

Added value: +{
+  "anyOf": [
+    {
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "default": null,
+  "description": "An agency client's email, from list_agency_clients, to run this call in that client's account: their agents, settings, connected accounts, and credits. Omit it to act on your own account."
+}

First observedSep 15, 2026

TDQS

A4.9/5.0

Behavior5/5

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

Annotations supply only readOnlyHint=true, and the description goes well beyond it: per-query credit cost with failed-query-free and capped soft-fail, exact-URL dedup with downstream cross-outlet dedup at `record_search_results`, server-side day/week/month/year recency bucketing that can exceed `lookback_days`, and distinct failure semantics (`google_news_available: False` for outages vs. genuinely empty, `InsufficientCreditsError` pre-run refusal). It also warns that results are raw articles, not company rows.

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 purpose, then query construction, then cost — the ordering is sensible and most sentences carry operational detail. There is some redundancy, e.g. URL dedup is stated in the summary and again in the returns block, and a few phrasings are padded, but nothing is filler at the level of the LOW examples.

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

Completeness5/5

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

No output schema exists, and the `<returns>` block fully specifies the row shape (`title, url, source, snippet, date, published_at, query`) plus the partial-vs-total failure distinction. Combined with cost, dedup, and recency-window caveats, an agent has everything needed to call it and interpret results.

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

Parameters5/5

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

Schema coverage is 67% and `lookback_days` has no schema description at all, so the description compensates by explaining its default of 30 and the coarse-bucket overrun caveat. It also adds query-construction semantics (journalist-headline keywords, one per signal) that the schema's brief `queries` note does not cover.

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?

States a specific verb and resource ('Discover companies from recent news via Google News keyword search') and explicitly positions itself among three named siblings (`search_news_events`, `tavily_search`, `exa_search_news`). It even names its differentiating edge — hyper-local and trade-press index coverage — so an agent can route without opening another schema.

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 signals (expansion stories, plant incidents, small-market exec moves) and explicit when-not/how-not guidance: semantic full-sentence phrasing returns ZERO results, and geographic scoping via keywords is brittle and should be filtered downstream. It also prescribes batching ('one query per signal, all in one call') and notes the News-mode skill owns routing.

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

All Sliq tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?