Skip to main content
Glama

veritas-8183-3star-rating-seal

audit_ai_inference_escrow

Idempotent

[Private] Were you overbilled or served a cheaper model? Checks LLM output for token inflation, repetition collapse and missing required terms. Example: billedOutputTokens=32.

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": {
      "responseText": "<responseText>"
    },
    "name": "audit_ai_inference_escrow"
  }
}

Input Schema

NameRequiredDescriptionDefault
responseTextYesGenerated text output returned by the inference provider.
forbiddenTermsNoTerms that must not appear in the output.
mustIncludeTermsNoKeywords or phrases that must appear in the output.
billedOutputTokensNoNumber of output tokens billed by the provider.
minUniqueWordRatioNoMinimum lexical diversity ratio (0 to 1) to detect model downgrade.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed6 schema fields changed
    • addedInput schema / properties / billedOutputTokens / description
      Added value: +"Number of output tokens billed by the provider."
    • addedInput schema / properties / forbiddenTerms / description
      Added value: +"Terms that must not appear in the output."
    • addedInput schema / properties / minUniqueWordRatio / description
      Added value: +"Minimum lexical diversity ratio (0 to 1) to detect model downgrade."
    • addedInput schema / properties / mustIncludeTerms / description
      Added value: +"Keywords or phrases that must appear in the output."
    • addedInput schema / properties / responseText / description
      Added value: +"Generated text output returned by the inference provider."
    • changedInput schema / required
      Previous value: -[
      -  "responseText",
      -  "billedOutputTokens"
      -]New value: +[
      +  "responseText"
      +]
  2. First observed

TDQS

B3.4/5.0
Behavior3/5

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

Annotations already declare idempotent, non-destructive, closed-world behavior, so the bar is lower. The description adds that it compares billed tokens against actually generated text, which is useful. It does not explain why the tool is not read-only (annotations say readOnlyHint=false) or what happens with escrow funds, leaving a real behavioral gap for a tool with 'escrow' in its name.

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?

Three short sentences, front-loaded with the value proposition and followed by the concrete checks. The '[Private]' tag and dangling example are terse but not wasteful; nothing repeats the name or title.

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

Completeness3/5

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

There are five parameters and no output schema, so the description should indicate what an audit returns (a verdict, flags, a refund amount). Instead it stops at listing what is inspected, leaving the return shape and the escrow consequence unexplained.

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 description coverage is 100%, so all five parameters are already documented in the schema, and the description only echoes billedOutputTokens via the 'Example: billedOutputTokens=32' fragment. It adds no insight into how billing versus observed tokens is compared or what minUniqueWordRatio thresholds mean in practice.

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 uses a concrete verb and resource ('Checks LLM output') and enumerates the specific conditions it detects: token inflation, repetition collapse and missing required terms. That is far more specific than a tautology, though it never positions itself against the audit_* escrow siblings that share its prefix.

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?

The opening question ('Were you overbilled or served a cheaper model?') implies the triggering situation, so usage is inferable. There is no explicit when-not guidance, no prerequisite for the private/escrow context the name implies, and no named alternative for related auditing needs.

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.