rag-quality-testing git:20261011.7d9502bA

Immutable. This exact content never changes and is served at /api/v1/blob/f1c8fc5d38cfa94e.

---
name: rag-quality-testing
description: Use this skill when you need evidence-bounded grounding, relevance, completeness, citation support, abstention, and answer-level evidence in RAG outputs; triggers include RAG 质量 and RAG quality.
---

# RAG Quality Testing

## When to Use

- Use this skill when you need evidence-bounded analysis, design, or validation preparation for grounding, relevance, completeness, citation support, abstention, and answer-level evidence in RAG outputs.
- Use it to review an Agent, RAG, or LLM plan, result, or evidence set and produce actionable improvements.
- Use it when context is incomplete but a bounded first pass with assumptions, gaps, and human-decision boundaries is still useful.

## Output Format Options

- Default to Markdown organized by domain risk, evidence state, priority, and boundary.
- When the user requests tables, CSV, JSON, or ticket fields, preserve the same finding fields, evidence, and decision boundaries.
- Before machine consumption, confirm the schema, enums, required fields, and evidence sources.

## How to Use

1. Read and follow `prompts/rag-quality-testing.md`, including its input audit, domain coverage, and output order.
2. Extract scope, environment, version, time window, constraints, success criteria, and available evidence, with attention to grounding, relevance, completeness, citation support, abstention.
3. Separate confirmed facts, evidence-backed inferences, candidate recommendations, and Human decisions before ranking by risk and evidence strength.
4. Turn high-risk items into preconditions, steps, expected behavior or decision criteria, required evidence, and a validation method.
5. When input is incomplete, deliver a bounded first pass, state unsupported conclusions, and never present static material as execution evidence.

## Reference Files

- When calibrating domain outcomes, valid controls, or evidence gaps, read [references/domain-decision-example.md](references/domain-decision-example.md) for sanitized input, per-item oracles, and reproducible calculations.

- Always read `prompts/rag-quality-testing.md`; it is the complete execution specification for this skill.
- For evaluation, read `evals/eval.yaml` and the matching cases under `evals/cases/`.
- Load `references/`, `examples/`, `scripts/`, or `output-formats.md` only when those directories exist and the task needs them.

## Domain Execution Essentials

1. Fix query/sample, knowledge/index/retrieval/model/prompt versions. Map queries to retrieved chunks, atomic claims and citations with support/contradiction/gap labels.
2. Judge citation accuracy, claim support and completeness separately. Supported claims may omit required conditions; use the task and project rubric.
3. Pair answerable in-corpus cases with out-of-corpus refusal controls. Assess task correctness, over-refusal and unsupported answers separately; refusal rate is no quality proxy.
4. Missing chunks, ground truth, permissions or gates remain UNKNOWN/NOT_RUN. Distinguish retrieval failure from generation misreading, retaining evidence boundaries and responsible stages.

## Core Constraints

- Keep the analysis focused on grounding, relevance, completeness, citation support, abstention, and answer-level evidence in RAG outputs; do not replace business owners or Human risk acceptance, exception approval, or safety decisions.
- Never invent system behavior, fields, model outputs, data, thresholds, root causes, execution records, or pass claims.
- Static design, plans, file presence, or a dry run retain their evidence state and cannot become proof of real execution.
- When evidence is insufficient, use pending confirmation, blocked, unassessed, or NOT_SCORED and give the smallest validation method.
- For user data, production, or safety work, use least privilege, masked data, mocks, dry runs, or isolation.

## Delivery Checklist

- [ ] Covered grounding, relevance, completeness, citation support, abstention, with source, evidence state, and validation method for each.
- [ ] Separated facts, inferences, candidate recommendations, gaps, and Human decisions.
- [ ] Gave high-risk items P0/P1/P2/P3 or an equivalent priority, owner role, and close condition.
- [ ] Did not turn plans, static checks, or dry runs into test execution, all-passed, or safety-approved claims.
- [ ] Stated residual risk, stop/escalation conditions, and next actions.

## Common Pitfalls

- Listing checks without triggers, expected concerns, owner roles, close conditions, and evidence.
- Treating adjacent tests or model tools as a complete RAG quality judgment.
- Using unexplained numbers for false precision or writing correlation as causation.
- Refusing incomplete input, or pretending that incomplete evidence is conclusive.

## Best Practices

- Start with paths most likely to cause user harm, business loss, or decision blockage.
- Use the smallest verifiable experiment to reduce uncertainty and record conditions, versions, sources, and evidence.
- Make the Skill independently installable, executable, and reviewable by another engineer.