Skip to main content
Glama

label eval pair

label_eval_pair

Record a human preference verdict for a candidate-baseline sample to build calibration evidence, producing a corrected win rate with uncertainty interval.

Instructions

Record a HUMAN's verdict on one candidate-vs-baseline pair of a comparison run — the calibration evidence behind the corrected win rate. A pairwise judge's printed win rate is compressed toward 50/50 (a true 80/20 prints ~70/30 even for a judge at the human ceiling); from 30 labels the run reports a corrected rate with an interval that carries the calibration uncertainty. POST /v1/evals/{id}/pair_labels (API-key scope: evals:write). Returns: {sample_index, candidate, verdict, critique, pairwise} — pairwise is the run's refreshed calibration block (same shape as get_eval_pairwise), so one call shows what the label bought. One label per (sample, candidate); posting again overwrites. DELETE /v1/evals/{id}/pair_labels?sample_index=…&candidate=… removes one; GET lists them. Notes: Labelling is a human's job: only relay verdicts the user actually gave — never invent preferences to reach 30. 400 with the offender named on a bad sample_index/candidate/verdict; 404 when the run is not in this workspace. No spend.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
idYesThe eval run id (a comparison run; criterion runs are refused — they are graded pass/fail per exchange with trace labels).
verdictYesWhich answer the human preferred: "candidate", "baseline", or "tie" (a tie is a real answer, not a skip).
critiqueNoOptional free-text WHY (≤2000 chars).
candidateYesThe candidate arm key the verdict is about (as listed in candidate_models / get_eval_samples).
sample_indexYes0-based sample index within the run (the order get_eval_samples returns).
Install Server

TDQS

A4.7/5.0
Behavior5/5

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

With annotations already providing readOnly/idempotent/destructive hints, the description adds substantial behavioral context: one-label-per-pair overwrite semantics, companion DELETE/GET endpoints, the returned pairwise calibration block, 400/404 error cases, API-key scope, 'No spend', and the integrity rule about not fabricating preferences.

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

Conciseness5/5

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

The description is dense but well-organized, front-loading the core purpose before covering endpoint details, return shape, mutation semantics, error handling, and constraints. Every sentence earns its place by communicating a distinct fact an agent needs to call this tool correctly.

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?

The tool is complex, yet the description covers selection criteria, request semantics, response contents, side effects, error handling, authentication scope, and cost. Since there is no output schema, the description compensates by explicitly describing the returned pairwise calibration block and linking it to get_eval_pairwise.

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% and each parameter already has a descriptive explanation, including the meaning of verdict values and sample_index ordering. The description adds contextual framing like candidate-vs-baseline and overwrite behavior, but it does not need to restate parameter formats, so the baseline of 3 applies.

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?

The description opens with a specific verb and resource: 'Record a HUMAN's verdict on one candidate-vs-baseline pair of a comparison run.' It clearly frames this as calibration evidence and differentiates it from read-side siblings like get_eval_pairwise by emphasizing the write action and human-labeling role.

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?

The description explicitly scopes usage to comparison runs, states that criterion runs are refused, and warns that labels must come from actual user verdicts, never invented. It also cites the 30-label corrected-rate threshold, giving the agent concrete conditions for when this tool should be invoked.

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

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/omnia-v/errorbar-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server