idea-creator

name	idea-creator
description	Prompt-only by default in the lightweight pack; Claude/Gemini reviewer transport is explicit opt-in. Generate and rank research ideas given a broad direction. Use when user says "找idea", "brainstorm ideas", "generate research ideas", "what can we work on", or wants to explore a research area for publishable directions.

Override for Codex users who want Gemini, not a second Codex agent, to act as the reviewer. Install this package after skills/skills-codex/*.

Customized Pack Defaults

This reviewer overlay is opt-in transport only. In the customized lightweight pack, default to prompt-only review unless the user explicitly asks to use this overlay reviewer backend:

• Write the review prompt under review-prompts/ and ask the user to run it in a separate conversation with the relevant skills enabled.
• Do not call Claude/Gemini reviewer MCP bridges, APIs, or follow-up status tools by default.
• Use pasted review text as the external verdict, then write compact summaries such as REVIEW_SUMMARY.md, NEXT_ACTIONS.md, or the skill-specific log.
• Keep heavy experiments AutoDL/HPC-gated and user-approved; local work stays at edits, tests, lint, dry-runs, and tiny smoke checks.

Research Idea Creator

Generate publishable research ideas for: $ARGUMENTS

Overview

Given a broad research direction from the user, systematically generate, validate, and rank concrete research ideas. This skill composes with /research-lit, /novelty-check, and /research-review to form a complete idea discovery pipeline.

Constants

• PILOT_MAX_HOURS = 2 — Skip any pilot estimated to take > 2 hours per GPU. Flag as "needs manual pilot".

• PILOT_TIMEOUT_HOURS = 3 — Hard timeout: kill pilots exceeding 3 hours. Collect partial results if available.

• MAX_PILOT_IDEAS = 3 — Pilot at most 3 ideas in parallel. Additional ideas are validated on paper only.

• MAX_TOTAL_GPU_HOURS = 8 — Total GPU budget for all pilots combined.

• REVIEWER_MODEL = gemini-review — Gemini reviewer invoked through the local gemini-review MCP bridge for brainstorming and critique. Set GEMINI_REVIEW_MODEL if you need a specific Gemini model override.

• OUTPUT_DIR = idea-stage/ — Directory for idea output files.

💡 Override via argument, e.g., /idea-creator "topic" — pilot budget: 4h per idea, 20h total.

Workflow

Phase 0: Load Reference-Led Topic Seeds (if available)

Before the landscape survey, check idea-stage/TOPIC_CANDIDATES.md. If it exists, read it as a compact evidence handoff. Consult idea-stage/REFERENCE_DECONSTRUCTION.md only when a locator or boundary needs interpretation; do not redo its paper or figure evidence extraction.

For every carried seed, preserve its candidate ID, gap source type, source locators, gate status, minimum discriminating evidence, and conclusion boundary. A ready_for_novelty_check seed is still unverified; needs_evidence becomes an explicit research task; blocked remains excluded unless a narrower repairable question is supported.

Phase 1: Landscape Survey (5-10 min)

Map the research area to understand what exists and where the gaps are.

When reference-led seeds exist, use their source locators as starting queries, then test each asserted gap against broader and more recent literature. Keep the seed's status and conclusion boundary visible in the landscape; do not turn a single paper's future-work statement or limitation into a field-wide claim.

1. Scan local paper library first: Check papers/ and literature/ in the project directory for existing PDFs. Read first 3 pages of relevant papers to build a baseline understanding before searching online. This avoids re-discovering what the user already knows.

2. Search recent literature using WebSearch:

   • Top venues in the last 2 years (NeurIPS, ICML, ICLR, ACL, EMNLP, etc.)
   • Recent arXiv preprints (last 6 months)
   • Use 5+ different query formulations
   • Read abstracts and introductions of the top 10-15 papers

3. Build a landscape map:

   • Group papers by sub-direction / approach
   • Identify what has been tried and what hasn't
   • Note recurring limitations mentioned in "Future Work" sections
   • Flag any open problems explicitly stated by multiple papers

4. Identify structural gaps:

   • Methods that work in domain A but haven't been tried in domain B
   • Contradictory findings between papers (opportunity for resolution)
   • Assumptions that everyone makes but nobody has tested
   • Scaling regimes that haven't been explored
   • Diagnostic questions that nobody has asked

Phase 2: Idea Generation (brainstorm with external LLM)

Use the local gemini-review MCP bridge for divergent thinking:

mcp__gemini-review__review_start:
  prompt: |
    You are a senior ML researcher brainstorming research ideas.

    Research direction: [user's direction]

    Here is the current landscape:
    [paste landscape map from Phase 1]

    Key gaps identified:
    [paste gaps from Phase 1]

    Reference-led topic seeds (optional):
    [paste candidate ID, source locators, gate status, minimum discriminating
    evidence, and conclusion boundary from TOPIC_CANDIDATES.md]

    Treat a seed as a constrained starting point, not a novelty verdict. For
    every idea descended from a seed, preserve its provenance and do not claim
    more than its recorded conclusion boundary. `needs_evidence` means the
    missing evidence must appear in the validation plan; `blocked` is not a
    candidate to promote without a narrower, evidence-backed repair.

    Generate 8-12 concrete research ideas. For each idea:
    1. One-sentence summary
    2. Core hypothesis (what you expect to find and why)
    3. Minimum viable experiment (what's the cheapest way to test this?)
    4. Expected contribution type: empirical finding / new method / theoretical result / diagnostic
    5. Risk level: LOW (likely works) / MEDIUM (50-50) / HIGH (speculative)
    6. Estimated effort: days / weeks / months

    Prioritize ideas that are:
    - Testable with moderate compute (8x RTX 3090 or less)
    - Likely to produce a clear positive OR negative result (both are publishable)
    - Not "apply X to Y" unless the application reveals genuinely surprising insights
    - Differentiated from the 10-15 papers above

    Be creative but grounded. A great idea is one where the answer matters regardless of which way it goes.

After this start call, immediately save the returned jobId and poll mcp__gemini-review__review_status with a bounded waitSeconds until done=true. Treat the completed status payload's response as the brainstorm output, and save the completed threadId for follow-up critique in Phase 4.

Phase 3: Mechanical consolidation + objective feasibility gate

This phase does NOT judge idea quality, novelty, or impact — those are the job of the Phase-4 cross-model reviewer (a different model family). Dropping ideas here on a same-family novelty or impact call would pre-filter the reviewer's input with same-family judgment — the opposite of why debuffer uses a cross-model reviewer at all. Phase 3 only (a) clusters near-duplicate ideas and (b) drops ideas that are OBJECTIVELY out of budget; everything else passes through ANNOTATED, not eliminated.

1. Objective feasibility gate (safe to gate here): drop an idea ONLY on a mechanical, budget-based fact — estimated compute > 1 week of available GPU time, OR a dataset that is provably unavailable. Do NOT drop on "implementation looks complex" — annotate complexity instead.

2. Novelty signal — ANNOTATE, do not eliminate: do 2-3 targeted searches and attach a prior_work note (what looks related, with links). This is input for the Phase-4 reviewer, not a filter; full /novelty-check runs in Phase 4. Do NOT drop an idea here because it "might already be done."

3. Impact signal — ANNOTATE, do not eliminate: attach a one-line so_what note (why the result would matter either way). Do NOT drop on a same-family "a reviewer wouldn't care" call — that is exactly what the Phase-4 cross-model reviewer is for.

Every feasible, non-duplicate idea — with its prior_work and so_what annotations — proceeds to Phase 4, where the cross-model reviewer does the quality/novelty narrowing.

Phase 4: Deep Validation (for top ideas)

For each surviving idea, run a deeper evaluation:

1. Novelty check: Use the /novelty-check workflow (multi-source search + Gemini cross-verification) for each idea

2. Critical review: Use mcp__gemini-review__review_reply_start with the saved completed threadId:

   mcp__gemini-review__review_reply_start:
     threadId: [saved completed threadId from Phase 2]
     prompt: |
       Here are our top ideas after filtering:
       [paste surviving ideas with novelty check results]
   
       For each, play devil's advocate:
       - What's the strongest objection a reviewer would raise?
       - What's the most likely failure mode?
       - How would you rank these for a top venue submission?
       - Which 2-3 would you actually work on?
   

   After this start call, immediately save the returned jobId and poll mcp__gemini-review__review_status with a bounded waitSeconds until done=true. Treat the completed status payload's response as the follow-up critique.

3. Combine rankings: Merge your assessment with Gemini's ranking. Select top 2-3 ideas for pilot experiments.

Phase 5: Parallel Pilot Experiments (for top 2-3 ideas)

Before committing to a full research effort, run cheap pilot experiments to get empirical signal. This is the key differentiator from paper-only validation.

1. Design pilots: For each top idea, define the minimal experiment that would give a positive or negative signal:

   • Single seed, small scale (e.g., small dataset subset, fewer epochs)
   • Target: 30 min - PILOT_MAX_HOURS per pilot on 1 GPU
   • Estimate GPU-hours BEFORE launching. If estimated time > PILOT_MAX_HOURS, reduce scale (fewer epochs, smaller subset) or flag as "needs manual pilot"
   • Clear success metric defined upfront (e.g., "if metric improves by > 1%, signal is positive")

2. Deploy in parallel: Use /run-experiment to launch pilots on different GPUs simultaneously:

   GPU 0: Pilot for Idea 1
   GPU 1: Pilot for Idea 2
   GPU 2: Pilot for Idea 3
   

   Use run_in_background: true to launch all at once.

3. Collect results: Use /monitor-experiment to check progress. If any pilot exceeds PILOT_TIMEOUT_HOURS, kill it and collect partial results. Once all pilots complete (or timeout), compare:

   • Which ideas showed positive signal?
   • Which showed null/negative results? (eliminate or deprioritize)
   • Any surprising findings that suggest a pivot?
   • Total GPU-hours consumed (track against MAX_TOTAL_GPU_HOURS budget)

4. Re-rank based on empirical evidence: Update the idea ranking using pilot results. An idea with strong pilot signal jumps ahead of a theoretically appealing but untested idea.

Note: Skip this phase if the ideas are purely theoretical or if no GPU is available. Flag skipped ideas as "needs pilot validation" in the report.

Phase 6: Output — Ranked Idea Report

Write a structured report to idea-stage/IDEA_REPORT.md:

# Research Idea Report

**Direction**: [user's research direction]
**Generated**: [date]
**Ideas evaluated**: X generated → Y survived filtering → Z piloted → W recommended

## Landscape Summary
[3-5 paragraphs on the current state of the field]

## Recommended Ideas (ranked)

### Idea 1: [title]
- **Hypothesis**: [one sentence]
- **Reference-led provenance**: [candidate ID + source locators, or none]
- **Evidence gate**: [source status + evidence still required]
- **Conclusion boundary**: [strongest supported claim and excluded claim]
- **Minimum experiment**: [concrete description]
- **Expected outcome**: [what success/failure looks like]
- **Novelty**: X/10 — closest work: [paper]
- **Feasibility**: [compute, data, implementation estimates]
- **Risk**: LOW/MEDIUM/HIGH
- **Contribution type**: empirical / method / theory / diagnostic
- **Pilot result**: [POSITIVE: metric +X% / NEGATIVE: no signal / SKIPPED: needs GPU]
- **Reviewer's likely objection**: [strongest counterargument]
- **Why we should do this**: [1-2 sentences]

### Idea 2: [title]
...

## Eliminated Ideas (for reference)
| Idea | Reason eliminated |
|------|-------------------|
| ... | Already done by [paper] |
| ... | Requires > 1 week GPU time |
| ... | Result wouldn't be interesting either way |

## Pilot Experiment Results
| Idea | GPU | Time | Key Metric | Signal |
|------|-----|------|------------|--------|
| Idea 1 | GPU 0 | 45 min | +2.3% CE | POSITIVE |
| Idea 2 | GPU 1 | 30 min | -0.1% CE | NEGATIVE |
| Idea 3 | GPU 2 | 1.5 hr | +0.8% CE | WEAK POSITIVE |

## Suggested Execution Order
1. Start with Idea 1 (positive pilot signal, lowest risk)
2. Idea 3 as backup (weak signal, may need larger scale to confirm)
3. Idea 2 eliminated by pilot — negative result documented

## Next Steps
- [ ] Scale up Idea 1 to full experiment (multi-seed, full dataset)
- [ ] If confirmed, invoke /auto-review-loop for full iteration

Output Protocols

Follow these shared protocols for all output files:

• Output Versioning Protocol — write timestamped file first, then copy to fixed name
• Output Manifest Protocol — log every output to docs/project/OUTPUT_MANIFEST.md
• Output Language Protocol — respect the project's language setting

Key Rules

• Large file handling: If the Write tool fails due to file size, immediately retry using Bash (cat << 'EOF' > file) to write in chunks. Do NOT ask the user for permission — just do it silently.

• The user provides a DIRECTION, not an idea. Your job is to generate the ideas.

• Quantity first, quality second: brainstorm broadly, then filter ruthlessly.

• A good negative result is just as publishable as a positive one. Prioritize ideas where the answer matters regardless of direction.

• Don't fall in love with any idea before validating it. Be willing to kill ideas.

• Always estimate compute cost. An idea that needs 1000 GPU-hours is not actionable for most researchers.

• "Apply X to Y" is the lowest form of research idea. Push for deeper questions.

• Include eliminated ideas in the report — they save future time by documenting dead ends.

• If the user's direction is too broad (e.g., "NLP", "computer vision", "reinforcement learning"), STOP and ask them to narrow it. A good direction is 1-2 sentences specifying the problem, domain, and constraint — e.g., "factorized gap in discrete diffusion LMs" or "sample efficiency of offline RL with image observations". Without sufficient specificity, generated ideas will be too vague to run experiments on.

Composing with Other Skills

After this skill produces the ranked report:

/idea-creator "direction"     → ranked ideas
/novelty-check "top idea"     → deep novelty verification (already done in Phase 4, but user can re-run)
/research-review "top idea"   → external critical feedback
implement                     → write code
/run-experiment               → deploy to GPU
/auto-review-loop             → iterate until submission-ready