Skip to main content
Glama
Sign In

Agent Data API

by io.github.Bustlll
OverviewDiscussionsAdmin

Web search (optionally read the top pages)

web_search
Read-onlyIdempotent

Search the web: Google-quality results re-ranked for authority, merged with papers (arXiv), GitHub and news when the query needs them; optional category. Use it to find pages; with read=N it also returns the full clean text of the top N pages; with research=true it reads up to 8 pages and extracts the passages that answer the query, with sources. For one known URL use web_read instead. Price: results only $0.006, read 1-5 pages $0.015, research $0.05. Returns titles, URLs, snippets (and page markdown with a prompt-injection verdict per page). Live results; 1-15 s. Paid with x402 (USDC on Base, Polygon or Solana) per call; failed or empty calls are never charged.

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": {
      "q": "<q>"
    },
    "name": "web_search"
  }
}

Input Schema

NameRequiredDescriptionDefault
qYesThe query, e.g. 'best espresso machine 2026'.
readNoPages to read in full (0 = results only, default 0).
timeNoOnly results from this period.
limitNoResults to return in results-only mode (default 10).
regionNoCountry code for results, e.g. us, uk, de, es (default us).
categoryNoOptional: favor one kind of source (results-only mode).
researchNotrue = research mode: read up to 8 pages and return key passages ($0.05).
max_charsNoMax characters per page read (default 4000).

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • addedInput schema / properties / category
      Added value: +{
+  "description": "Optional: favor one kind of source (results-only mode).",
+  "enum": [
+    "news",
+    "research paper",
+    "github",
+    "financial",
+    "company",
+    "pdf",
+    "docs"
+  ],
+  "type": "string"
+}
  2. Changed8 schema fields changed
    • addedInput schema / properties / limit / description
      Added value: +"Results to return in results-only mode (default 10)."
    • addedInput schema / properties / max_chars
      Added value: +{
+  "description": "Max characters per page read (default 4000).",
+  "maximum": 20000,
+  "minimum": 500,
+  "type": "integer"
+}
    • changedInput schema / properties / q / description
      Previous value: -"Search query"New value: +"The query, e.g. 'best espresso machine 2026'."
    • addedInput schema / properties / read
      Added value: +{
+  "description": "Pages to read in full (0 = results only, default 0).",
+  "maximum": 8,
+  "minimum": 0,
+  "type": "integer"
+}
    • changedInput schema / properties / region / description
      Previous value: -"us, uk, de, fr, es, ..."New value: +"Country code for results, e.g. us, uk, de, es (default us)."
    • addedInput schema / properties / research
      Added value: +{
+  "description": "true = research mode: read up to 8 pages and return key passages ($0.05).",
+  "type": "boolean"
+}
    • changedInput schema / properties / time / description
      Previous value: -"day, week, month or year"New value: +"Only results from this period."
    • addedInput schema / properties / time / enum
      Added value: +[
+  "day",
+  "week",
+  "month",
+  "year"
+]
  3. First observed

TDQS

A4.4/5.0
Behavior3/5

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

Annotations already establish readOnly, idempotent and openWorld, so the safety profile is covered. The description nonetheless adds genuinely non-obvious behavior: per-mode pricing ($0.006 / $0.015 / $0.05), x402 USDC payment on named chains, a 1-15 s latency band, a no-charge guarantee for failed/empty calls, and a prompt-injection verdict per page. Return contents (titles, URLs, snippets, markdown) are also disclosed; the only residual gap is pagination/result-count behavior across modes.

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 modes, then the sibling exclusion, then price/latency/payment. Nearly every clause earns its place, but the pricing and payment details are packed into dense semicolon-separated runs that could be trimmed without losing agent-relevant information.

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?

With eight parameters, two enums, mixed result modes and no output schema, the description is complete: it covers mode selection, cost, latency, payment mechanics, and the return payload including the prompt-injection verdict. An agent has everything needed to call it correctly on the first attempt.

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 coverage is 100%, so the baseline is 3; the description goes further by tying the read, research and category parameters to cost and output shape (results-only vs full text vs extracted passages with sources). It does not clarify the read/research interaction when both are set, nor max_chars/time/region, but those are fully documented in the schema.

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 ("Search the web") and immediately qualifies the result quality (authority re-ranked, merged with arXiv/GitHub/news). It explicitly names the sibling boundary with "For one known URL use web_read instead," so an agent can pick between the two without opening a 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 an explicit use case ("Use it to find pages") plus the three selection conditions: default results-only, read=N for full text of top N pages, research=true for passage extraction across up to 8 pages. Also supplies the exclusion rule routing single-URL lookups to web_read.

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

All Agent Data API toolsInspect
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