Clean room framing options

Clean-Room Framing Options

Purpose

Generate and stress-test six evidence-specific research framings from an evidence pack before a project story is chosen.

When To Use

Use before selecting a paper, report, proposal, or project narrative. It fits cases where the evidence may support several stories, where prior labels may bias interpretation, or where the obvious headline may be under-earned.

Required Inputs

• {{PROJECT_NAME}}
• {{RESEARCH_GOAL}}
• {{DOMAIN}}
• {{EVIDENCE_PACK}}
• {{TARGET_AUDIENCE}}
• {{AS_OF_DATE}}

Optional Inputs

• {{CURRENT_FRAMING}}: existing story to critique.
• {{LITERATURE_CONTEXT}}: sources or summaries that may affect novelty and positioning.
• {{CONSTRAINTS}}: what evidence can realistically be added later.
• {{SHARED_RESEARCH_CONTRACT}} for shared definitions covering authority order, claim statuses, provenance, and supersession.
• {{OUTPUT_LOCATION}}

Authority Order

Use {{SHARED_RESEARCH_CONTRACT}} when supplied. Otherwise apply this order by claim type:

1. Explicit user instructions, release constraints, safety constraints, and local project conventions.
2. Raw or near-raw project evidence for project claims.
3. Primary external sources for external claims.
4. Reproducible analyses and transparent secondary sources.
5. Canonical current-state project documents and authority ledgers.
6. Dated archival records and historical notes.
7. Input reports, model outputs, comments, and prior narratives, treated as leads.
8. Analyst or model judgment, labeled as judgment.

Tool And File-Access Assumptions

Use file access or browsing when available. If only the supplied text is available, stay inside that evidence boundary and state the limitation. If artifact creation is unavailable, output the full report in Markdown.

Handoffs

Upstream artifacts are an evidence pack, accepted literature synthesis, report arbitration output, pipeline evidence register, manuscript audit, or current framing.

Downstream artifacts are Six Candidate Framings, Claim Hygiene, Hostile Self-Critique, and What Would Most Change The Conclusion.

Pass the selected framing, live alternatives, and claim-hygiene notes to research-manuscript-audit, literature-review, or a writing refinement pass after the project story is chosen.

Process

1. Start from the evidence pack.
2. Identify empirical patterns, nulls, harms, exceptions, caveats, missing controls, and missing transfer checks.
3. Generate exactly six candidate framings that differ in what they claim, what evidence they compress, and what they would require.
4. Stress-test each framing against the strongest alternative explanations.
5. Keep multiple live options visible when the evidence leaves the choice open.
6. Use {{CURRENT_FRAMING}} as one candidate to test when supplied.
7. Use {{TARGET_AUDIENCE}} to score decision relevance and communication risk.
8. Use {{AS_OF_DATE}} for novelty and freshness labels.
9. Separate direct evidence, inference, uncertainty, judgment, novelty hypotheses, and claims blocked by missing evidence.
10. Assign one claim-support status from the shared vocabulary.

Claim-support statuses:

• Supported
• PartiallySupported
• ReasonableInference
• Ambiguous
• Overstated
• Unsupported
• StaleOrSuperseded
• Unverified
• Opinion

Work privately. Expose concise rationale, evidence, uncertainty, and decisions.

Output Contract

Produce exactly these sections:

1. Six Candidate Framings

For each candidate include:

• Framing label
• One-sentence thesis
• Empirical pattern it compresses
• Why it is evidence-specific
• Why it matters for {{DOMAIN}}
• Evidence in favor
• Evidence against or limiting evidence
• Assumptions or interpretive moves required
• Relative edge over the other candidates
• Main blocker to preferring it now
• Possible literature-facing novelty claim, labeled as hypothesis unless verified
• Risk of sounding derivative or generic
• Current status: live, plausible but narrow, or tempting but under-earned

2. Claim Hygiene

List:

• strongest Supported claims;
• useful PartiallySupported or ReasonableInference claims;
• Ambiguous or Unverified claims needing validation before external use;
• Overstated, Unsupported, or StaleOrSuperseded claims to remove or repair.

3. Hostile Self-Critique

Give the three strongest objections a skeptical reviewer could raise from the evidence alone. For each, state why it bites, which framings it damages, severity, and the most honest containment or reframing.

4. What Would Most Change The Conclusion

Identify the smallest number of additional observations, analyses, or experiments that could materially change which framings remain live. For each, include what to collect, what result pattern would matter, why it is high ROI, and whether it resolves conceptual, measurement, or external-validity uncertainty.

Write to {{OUTPUT_LOCATION}} if possible. Otherwise output the full Markdown report.

Quality Gate

Before finalizing, verify:

• Exactly six candidate framings are provided.
• Generic framings are sharpened into evidence-specific claims or downgraded.
• Literature novelty is stated as fact only when directly verified.
• Multiple live options remain visible when evidence leaves the choice open.
• Dependencies on prior projects, venues, file paths, deadlines, and models are explicit.