Skip to main content
Glama
Sign In

Passage Data — cited environmental limits

by io.github.kedu21-beep
OverviewDiscussionsAdmin

Ask a site a question

ask_site
Idempotent

What do a project's own documents say about a question, cited to the page? A question answered from one site's own read documents, by Claude, and checked before it is shown: every statement carries quotes, each quote is found again on the page it names, and a statement whose quote or number is not on its page is cut and counted (n_cut). quotes give the document, its id and the page; reasoning is Claude's own and not from the documents; not_searched names the documents not read yet. Priced before any call and refused over $0.25 a question (ask.over_cap); an account asks at most 100 questions and spends at most $5 with Claude a UTC day (ask.over_ceiling, spend.daily_cap). cost says what the answer spent; the same question over the same pages is answered from memory at $0.

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": {
      "site": "<site>",
      "question": "<question>"
    },
    "name": "ask_site"
  }
}

Input Schema

NameRequiredDescriptionDefault
siteYesThe site: its id (from list_sites), or a part of its name that names one site only.
questionYesThe question, in any language; numbered parts (1. … 2. …) are answered part by part.
skip_pagesNoRead the next pages instead: the number of best-ranked pages an earlier answer already read (its `searched.read`).

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A3.8/5.0
Behavior5/5

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

With annotations only saying readOnlyHint=false and idempotentHint=true, the description carries the real behavioral load: answers are verified quote-by-quote, unverifiable statements are cut and counted (n_cut), costs are priced before the call, questions are refused over $0.25, and the account is capped at 100 questions/$5 per UTC day. It also discloses that repeated identical questions are served from memory, which explains the idempotency hint.

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

Conciseness3/5

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

The opening sentence front-loads the purpose well, but the body is a dense run-on that mixes return-field semantics, pricing, and quota rules without clear separation. Every sentence carries information, yet the packing and question-and-answer phrasing make it harder to parse than it needs to be.

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?

There is no output schema, and the description compensates by explaining the return shape (quotes with document/id/page, reasoning, not_searched, cost, n_cut), which is the right call. Cost, quota, and verification behavior are all covered; only the practical routing against search/fetch alternatives is missing.

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

Parameters3/5

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

Schema coverage is 100%, so site, question, and skip_pages are already documented in the schema, so the baseline of 3 applies. The description reinforces skip_pages by tying 'not_searched names the documents not read yet' to the paging behavior, but adds no syntax or format detail beyond the schema.

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

Purpose4/5

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

The description names a specific verb+resource: answering a question from one site's own read documents, with page-level citations and verification. That is clearly distinct from siblings like search, fetch, or verify_claim, even though no sibling is named explicitly. The question-style opener is stylistic but the substance is concrete.

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

Usage Guidelines3/5

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

Usage is only implied: the note that 'the same question over the same pages is answered from memory at $0' hints at when re-asking is cheap, and the skip_pages parameter implies continuing an earlier answer. There is no explicit when-to-use/when-not guidance or routing to alternatives such as search or fetch.

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

All Passage Data — cited environmental limits 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