@sophia-reedrag-knowledge-evaluation-redteamテキスト公開2026/06/14 更新

RAG Knowledge Ops prompt that builds an evaluation suite for high-risk AI workflows and returns eval matrix, adversarial cases, grading rubric, and release threshold.

62スター0Fork139コピー

Prompt

プレビュー

Claude で開く

# Role
You are a knowledge systems architect who designs retrieval workflows that cite and stay current. You know current AI product patterns: tool calling, structured outputs, long-context reasoning, multimodal generation, MCP connectors, evals, trace review, and human approval gates.

# Mission
Build an evaluation suite for high-risk AI workflows for RAG ingestion, retrieval quality, citation design, and knowledge base governance.

# Inputs
- Project context: a support answer engine grounded in product docs and release notes
- Target audience: AI engineers, knowledge managers, legal ops, support ops
- Success metric: activation, quality, revenue, risk reduction, or cycle time
- Available tools and data: vector database, document parser, reranker, eval set, access control list
- Constraints: source-grounded answers; freshness policy; explicit unknown handling
- Desired depth: Fast triage
- Output tone: Clear operator memo

# Domain signals to inspect
- source documents
- chunk samples
- failed queries
- citation audits

# Known risks to control
- stale retrieval
- citation mismatch
- overly broad chunks
- permission leakage

# Method
1. Convert user outcomes into measurable tasks
2. Include realistic failure modes
3. Separate correctness from style
4. Define pass, warn, and block thresholds

# Required output sections
1. Success criteria
2. Golden tasks
3. Adversarial tasks
4. Rubric
5. Sampling plan
6. Release decision

# Output contract
Return Markdown only.
Use concise headings, decision tables, and checklists.
Every recommendation must tie back to evidence, assumptions, or a missing-input note.

# Quality bar
- Make the output reusable by another practitioner without extra explanation.
- Separate facts, assumptions, and recommendations.
- Use variables and placeholders only where the user must fill in project-specific information.
- Ask at most 3 clarifying questions only if missing information would materially change the result.
- Do not invent private facts, metrics, laws, integrations, or user quotes.
- Prefer specific operational language over generic AI enthusiasm.
- Include a "Verification" subsection when the output could affect users, money, security, compliance, health, or public trust.
- If the task involves tools or automation, state which actions are read-only, which actions write data, and which actions require human approval.

生成物

1 個の生成物

Example Output: RAG Knowledge Ops Evaluation and Red-Team Harness

Inputs used

Project context: a support answer engine grounded in product docs and release notes
Target audience: AI engineers, knowledge managers, legal ops, support ops
Success metric: activation, quality, and risk reduction
Available tools and data: vector database, document parser, reranker, eval set, access control list
Desired depth: Production-ready
Output tone: Clear operator memo

Generated Result

eval matrix, adversarial cases, grading rubric, and release threshold

Success criteria

Create at least 12 golden tasks: 6 normal cases, 3 edge cases, and 3 adversarial cases targeting stale retrieval. A passing result must cite the evidence source and state confidence.

Golden tasks

Create at least 12 golden tasks: 6 normal cases, 3 edge cases, and 3 adversarial cases targeting citation mismatch. A passing result must cite the evidence source and state confidence.

Adversarial tasks

Use failed queries as evidence, apply the constraint "explicit unknown handling", and explicitly note how the plan reduces overly broad chunks. The output should be ready for a practitioner to act on without a follow-up explanation.

Rubric

Use citation audits as evidence, apply the constraint "source-grounded answers", and explicitly note how the plan reduces permission leakage. The output should be ready for a practitioner to act on without a follow-up explanation.

Sampling plan

Release in three gates: internal dry run, limited pilot, then measured expansion. Each gate must show evidence that freshness policy is true in practice, not only in documentation.

Release decision

Release in three gates: internal dry run, limited pilot, then measured expansion. Each gate must show evidence that explicit unknown handling is true in practice, not only in documentation.

Recommended Decision

Proceed with a narrow pilot focused on source documents and chunk samples. Treat stale retrieval as the primary launch blocker. The first milestone should prove that the workflow produces a usable RAG blueprint, source policy, and retrieval eval plan with clear evidence, named owners, and a review path for ambiguous cases.

Expected quality checks

The result is specific to RAG ingestion, retrieval quality, citation design, and knowledge base governance.
It includes the required sections: Success criteria, Golden tasks, Adversarial tasks, Rubric, Sampling plan, Release decision.
It separates evidence, assumptions, risks, and recommended next actions.
It includes practical verification steps, not only generic advice.
It names the most important failure mode for this domain: stale retrieval.

Reuse note

Before copying the output into production work, replace all default variables with your real data and run a human review for high-impact decisions.

README

README.md

RAG Knowledge Ops: Evaluation and Red-Team Harness

Use this prompt when you need eval matrix, adversarial cases, grading rubric, and release threshold for RAG ingestion, retrieval quality, citation design, and knowledge base governance.

Best for

AI engineers, knowledge managers, legal ops, support ops
Teams that already have partial context but need a sharper, reusable artifact
AI workflows where the output must be auditable, editable, and easy to hand off

How to use

Replace the variables in the prompt with your real project context.
Keep the default constraints unless your team has stronger internal rules.
Review the generated output against the checklist in the example artifact.

Design notes

This seed follows current prompting practice: explicit role, structured inputs, domain evidence, operational guardrails, and a concrete output contract. It is written in English for international PromptHub users.