SkillVaultskills Browse all 500 skills

AI Engineering · Version 1.0.0 · Reviewed 2026-08-02

Retrieval Evaluation Designer

Make AI behavior measurable and safer for labeled set construction and recall measurement with evidence, explicit trade-offs, and a verification plan.

4 method steps 6 documented failure modes 5 diagnostic checks 7 quality gates

Builds retrieval-specific evaluation so recall problems are not mistaken for generation problems.

₹149 one-time

Get this skill archive

What this skill helps you do

  • Labeled set construction
  • Recall measurement
  • Failure classification

How Retrieval Evaluation Designer works

You provide

Prompts, model versions, evaluation data, and failures

It inspects

Failure class and context sufficiency for labeled set construction

It decides

A recall measurement change with one variable moved

You verify

Pass rate per case class against a pinned baseline

What it checks first

Retrieval Evaluation Designer builds retrieval-specific evaluation so recall problems are not mistaken for generation problems. Use it when the work involves Labeled set construction, Recall measurement, Failure classification.

  1. Whether the failure is systematic across a class of inputs or random, which separates a capability gap from a sampling issue.
  2. Whether evaluation data overlaps training or prompt-development data, which invalidates the measurement.
  3. Token distribution of inputs and outputs, since cost and latency are driven by the tail, not the mean.
  4. Whether the system has a defined behavior for low confidence, or always produces an answer.
  5. Version pinning across model, prompt, retrieval, and tools, because an unpinned component makes regressions unattributable.

Failure modes it recognizes

  • Silent quality regression after a provider updates a model behind an unversioned alias.
  • Evaluation overfitting where the prompt was tuned on the same examples used to score it.
  • Cost and latency dominated by a small number of very long inputs that were never in the test set.
  • Tool-calling loops where the model retries a failing tool without a bounded attempt budget.
  • Confident fabrication when context is insufficient because no refusal path was defined.
  • Distribution shift where production inputs diverge from the evaluation set over time.

Answers it will reject

  • Judging quality by reading a few outputs, which cannot detect a regression of a few percent.
  • Using a larger model to fix a problem caused by missing context, paying more for the same failure.
  • Fine-tuning before exhausting prompting and retrieval, which is slower to iterate and harder to reverse.
  • Using an LLM judge without validating the judge against human labels on the same rubric.

Decision rules it applies

  • Establish a labeled evaluation set and a baseline before changing anything; without a baseline there is no improvement, only change.
  • Pin every version and change one component at a time.
  • Define and test the refusal path explicitly; a system that cannot say "I do not know" will fabricate.
  • Budget latency and cost on p95 token counts, not averages.

Evidence it asks for

  • Score per input class (easy, hard, adversarial, no-answer) so aggregate scores cannot hide a broken class.
  • Log model version, prompt version, and retrieval version on every request for regression attribution.
  • Track p50 and p95 tokens and cost per successful task, not per call.

The method inside

  1. Translate labeled set construction into observable risks and falsifiable acceptance criteria.
  2. Choose the cheapest test level that can expose failures in recall measurement.
  3. Add representative positive, negative, boundary, and regression cases for failure classification.
  4. Define deterministic pass/fail signals, ownership, and the release decision when a check fails.

Deliverables

  • Labeled set construction assessment
  • Recall measurement decision and action plan
  • Failure classification verification checklist

Evidence requirements

  • Prompts, model/version, tools, retrieval path, and examples
  • Evaluation dataset and failure cases
  • Latency, cost, privacy, and policy constraints

Quality gates

  • Every material claim traces to supplied evidence or is labeled as a hypothesis.
  • The response follows the declared deliverable contract.
  • No execution, access, measurement, or verification is invented.
  • Secrets and personal data are redacted rather than repeated.
  • The user receives a concrete independent verification step.
  • The relevant failure modes in this domain were considered rather than only the reported symptom.
  • No listed anti-pattern was recommended as a solution.

Example task

Input

Our RAG answers are inconsistent and we cannot tell whether the retriever or the model is at fault.

Expected output

Separate them with a labeled query-to-document set and measure recall@k independently of any generation. If the correct passage is absent from the candidates, no prompt change can fix the answer, and that single measurement ends most of these debates...

Boundaries and compatibility

Ideal for

  • Labeled set construction: produce a decision or artifact grounded in supplied evidence.
  • Recall measurement: produce a decision or artifact grounded in supplied evidence.
  • Failure classification: produce a decision or artifact grounded in supplied evidence.

Out of scope

  • Treating prompt text as a security boundary
  • Claiming model quality from a handful of demos

Agent compatibility

  • GitHub Copilot custom agents
  • Claude Agent Skills / SKILL.md
  • Any instruction-following chat model

Tool policy: Advisory by default. No tools are assumed. If the host provides tools, use read-only evidence gathering unless the user explicitly approves a scoped write or execution action.