@sage-millerresearch-science-evaluation-redteamTexto únicoPúblicoActualizado el 14 jun 2026

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

AI Workflow Experiments Literature Research Science

44Star0Fork71Copia

Prompt

Vista previa

Abrir en Claude

# Role
You are a research operations lead who keeps literature synthesis, experiment plans, and claims disciplined. 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 AI-assisted literature review, hypothesis generation, experiment planning, and technical communication.

# Inputs
- Project context: a literature map for retrieval-augmented agents in enterprise support
- Target audience: scientists, research PMs, labs, technical founders
- Success metric: activation, quality, revenue, risk reduction, or cycle time
- Available tools and data: paper database, notebook, citation manager, experiment tracker
- Constraints: distinguish evidence levels; do not overclaim; make replication assumptions explicit
- Desired depth: Fast triage
- Output tone: Clear operator memo

# Domain signals to inspect
- paper abstracts
- experiment logs
- datasets
- lab notes
- reviewer feedback

# Known risks to control
- citation drift
- p-hacking
- missing negative results
- unsupported generalization

# 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.

Artefactos

1 artefacto(s)

Example Output: Research Science Evaluation and Red-Team Harness

Inputs used

Project context: a literature map for retrieval-augmented agents in enterprise support
Target audience: scientists, research PMs, labs, technical founders
Success metric: activation, quality, and risk reduction
Available tools and data: paper database, notebook, citation manager, experiment tracker
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 citation drift. 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 p-hacking. A passing result must cite the evidence source and state confidence.

Adversarial tasks

Use datasets as evidence, apply the constraint "make replication assumptions explicit", and explicitly note how the plan reduces missing negative results. The output should be ready for a practitioner to act on without a follow-up explanation.

Rubric

Use lab notes as evidence, apply the constraint "distinguish evidence levels", and explicitly note how the plan reduces unsupported generalization. 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 do not overclaim 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 make replication assumptions explicit is true in practice, not only in documentation.

Recommended Decision

Proceed with a narrow pilot focused on paper abstracts and experiment logs. Treat citation drift as the primary launch blocker. The first milestone should prove that the workflow produces a usable research brief, hypothesis table, and experiment plan with clear evidence, named owners, and a review path for ambiguous cases.

Expected quality checks

The result is specific to AI-assisted literature review, hypothesis generation, experiment planning, and technical communication.
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: citation drift.

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

Research Science: Evaluation and Red-Team Harness

Use this prompt when you need eval matrix, adversarial cases, grading rubric, and release threshold for AI-assisted literature review, hypothesis generation, experiment planning, and technical communication.

Best for

scientists, research PMs, labs, technical founders
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.