idea-discovery

Workflow 1: Idea Discovery Pipeline

Orchestrate a complete idea discovery workflow for: $ARGUMENTS

Overview

This skill chains sub-skills into a single automated pipeline:

/research-lit → /idea-creator → /novelty-check → /research-review → /research-refine-pipeline
  (survey)      (brainstorm)    (verify novel)    (critical feedback)  (refine method + plan experiments)

Each phase builds on the previous one's output. The final deliverables are a validated idea-stage/IDEA_REPORT.md with ranked ideas, plus a refined proposal (refine-logs/FINAL_PROPOSAL.md) and experiment plan (refine-logs/EXPERIMENT_PLAN.md) for the top idea.

Constants

• PILOT_MAX_HOURS = 2 — Skip any pilot experiment estimated to take > 2 hours per GPU. Flag as "needs manual pilot" in the report.
• PILOT_TIMEOUT_HOURS = 3 — Hard timeout: kill any running pilot that exceeds 3 hours. Collect partial results if available.
• MAX_PILOT_IDEAS = 3 — Run pilots for at most 3 top ideas in parallel. Additional ideas are validated on paper only.
• MAX_TOTAL_GPU_HOURS = 8 — Total GPU budget across all pilots. If exceeded, skip remaining pilots and note in report.
• AUTO_PROCEED = true — When true, checkpoints are informational: report the selected option and continue in the same turn. Set to false to ask for explicit user confirmation and end the turn at each selection checkpoint.
• REVIEWER_MODEL = gpt-5.6-sol — Model used via Codex MCP. Must be an OpenAI model (e.g., gpt-5.6-sol, o3, gpt-4o). Passed to sub-skills.
• OUTPUT_DIR = idea-stage/ — All idea-stage outputs go here. Create the directory if it doesn't exist.
• ARXIV_DOWNLOAD = false — When true, /research-lit downloads the top relevant arXiv PDFs during Phase 1. When false (default), only fetches metadata. Passed through to /research-lit.
• COMPACT = false — When true, generate compact summary files for short-context models and session recovery. Writes idea-stage/IDEA_CANDIDATES.md (top 3-5 ideas only) at the end of this workflow. Downstream skills read this instead of the full idea-stage/IDEA_REPORT.md.
• RENDER_HTML = true — When true (default), auto-render idea-stage/IDEA_REPORT.md to HTML at workflow end via /render-html. Uses --no-review (the source MD already went through novelty + cross-model review during Phase 3). Set false to skip, or pass — render html: false.
• REF_PAPER = false — Reference paper to base ideas on. Accepts: local PDF path, arXiv URL, or any paper URL. When set, the paper is summarized first (idea-stage/REF_PAPER_SUMMARY.md), then idea generation uses it as context. Combine with base repo for "improve this paper with this codebase" workflows.
• RESUMABLE = true — Record stage evidence under .aris/runs/<run_id>.json and require a deterministic evidence gate before declaring the final report complete.

💡 These are defaults. Override by telling the skill, e.g., /idea-discovery "topic" — ref paper: https://arxiv.org/abs/2406.04329 or /idea-discovery "topic" — compact: true.

Checkpoint execution rule

Resolve AUTO_PROCEED once from $ARGUMENTS before Phase 0 and keep that mode for the entire workflow.

• AUTO_PROCEED=true is non-blocking. A checkpoint is a progress update, not a question. State the result and the automatically selected next action, then continue executing in the same turn. Do not ask for confirmation, request user input, sleep, wait for silence, or end the turn at a checkpoint.
• AUTO_PROCEED=false is blocking. Present the options, ask the user, and end the turn. Resume only after an explicit reply.

Never implement auto-proceed as “ask, then continue if there is no response.” Once a turn ends, silence cannot resume the workflow. The user can still interrupt a non-blocking run at any time.

This rule governs only AUTO_PROCEED-controlled selection checkpoints. If the user explicitly enables a Feishu interactive gate, that external approval or reply is an intentional blocking exception; wait for that user-controlled gate rather than treating it as a silence timeout. Feishu off/push-only modes remain non-blocking under AUTO_PROCEED=true.

Per-stage evidence gate (RESUMABLE = true)

Resolve run_state.py and idea_discovery_gate.py through the same canonical helper chain used by /research-pipeline: .aris/tools/ → tools/ → $ARIS_REPO/tools/ → ~/.aris/repo/tools/. If either helper is unavailable, the final report is BLOCKED; do not silently continue without a state record.

For a new run, derive <run_id> from the direction slug and date, then start this ordered state record with --executor <actual-Claude-model> (for example, claude-sonnet-4.5):

research-lit,idea-creator,novelty-check,research-review,research-refine-pipeline

For each phase, mark running on entry and done --artifact <path> only after its artifact is present. Use these artifact locators so the final gate can check the canonical report rather than scattered scratch files:

novelty-check and research-review are reviewer-bearing phases. A done status or a heading alone is not review evidence. After each phase has folded substantive findings into its anchored report section, first record it done, then, only after the configured reviewer actually returns a positive, identity-bearing verdict, record the cross-family receipt using the actual returned model and durable thread/trace id:

<resolved-python> <resolved-run_state.py> accept . <run_id> novelty-check --verdict-id "<thread-or-trace-id>" --reviewer "<actual-reviewer-model>"
<resolved-python> <resolved-run_state.py> accept . <run_id> research-review --verdict-id "<thread-or-trace-id>" --reviewer "<actual-reviewer-model>"

Never invent either value and never call accept without the positive verdict required by the run-state contract. For novelty-check, both PROCEED and PROCEED WITH CAUTION are positive verdicts — caution is guidance for the pilot, not a rejection; only ABANDON is negative. A negative verdict does not grant a review receipt. Leave the phase done and the final gate BLOCKED, select a surviving or new idea, then re-run that reviewer-bearing phase. Do the same if the reviewer is unavailable, returns no valid identity/response, or its output was not folded into the report.

At the end of Phase 5, run:

<resolved-python> <resolved-idea_discovery_gate.py> . <run_id> --report idea-stage/IDEA_REPORT.md

The gate writes its result to gates.idea-discovery-evidence in the run state. On PASS, it has validated (but never created) the two review receipts, all required artifacts, and non-empty anchored report sections. Per-phase acceptance stays with each stage's own cross-model gate. On a non-zero exit, it writes explicit BLOCKED: <stage> evidence missing lines to the report; do not present the workflow as complete. On — resume <run_id>, start from the first non-terminal phase and re-run the gate before finalizing.

Pipeline

Phase 0: Load Research Brief (if available)

Before starting any other phase, check for a detailed research brief in the project:

1. Look for RESEARCH_BRIEF.md in the project root (or path passed as $ARGUMENTS)
2. If found, read it and extract:
   • Problem statement and context
   • Constraints (compute, data, timeline, venue)
   • What the user already tried / what didn't work
   • Domain knowledge and non-goals
   • Existing results (if any)
3. Use this as the primary context for all subsequent phases — it replaces the one-line prompt
4. If both RESEARCH_BRIEF.md and a one-line $ARGUMENTS exist, merge them (brief takes priority for details, argument sets the direction)

If no brief exists, proceed normally with $ARGUMENTS as the research direction.

💡 Create a brief from the template: cp templates/RESEARCH_BRIEF_TEMPLATE.md RESEARCH_BRIEF.md — keep it to ~1-2 pages (4-8k chars); long material goes in separate files referenced by path.

Phase 0.5: Reference Paper Summary (when REF_PAPER is set)

Skip entirely if REF_PAPER is false.

Summarize the reference paper before searching the literature:

1. If arXiv URL (e.g., https://arxiv.org/abs/2406.04329):

   • Invoke /arxiv "ARXIV_ID" — download to fetch the PDF
   • Read the first 5 pages (title, abstract, intro, method overview)

2. If local PDF path (e.g., papers/reference.pdf):

   • Read the PDF directly (first 5 pages)

3. If other URL:

   • Fetch and extract content via WebFetch

4. Generate idea-stage/REF_PAPER_SUMMARY.md:

# Reference Paper Summary

**Title**: [paper title]
**Authors**: [authors]
**Venue**: [venue, year]

## What They Did
[2-3 sentences: core method and contribution]

## Key Results
[Main quantitative findings]

## Limitations & Open Questions
[What the paper didn't solve, acknowledged weaknesses, future work suggestions]

## Potential Improvement Directions
[Based on the limitations, what could be improved or extended?]

## Codebase
[If `base repo` is also set: link to the repo and note which parts correspond to the paper]

🚦 Checkpoint: Present the summary to the user:

📄 Reference paper summarized:
- Title: [title]
- Key limitation: [main gap]
- Improvement directions: [2-3 bullets]

Proceeding to literature survey with this as context.

Phase 1 and Phase 2 will use idea-stage/REF_PAPER_SUMMARY.md as additional context — /research-lit searches for related and competing work, /idea-creator generates ideas that build on or improve the reference paper.

Phase 1: Literature Survey

Invoke /research-lit to map the research landscape. Idea discovery is exactly the place where Gemini's AI-driven broad coverage adds value, so include gemini as a source by default unless the user already specified an explicit — sources: directive in their idea-discovery invocation:

# If $ARGUMENTS already contains "— sources:", pass through unchanged
# (the user is in control of source selection):
/research-lit "$ARGUMENTS" — composed: idea-stage/IDEA_REPORT.md

# Otherwise (the common case), include gemini explicitly for broader discovery:
/research-lit "$ARGUMENTS" — sources: all, gemini — composed: idea-stage/IDEA_REPORT.md

— composed: idea-stage/IDEA_REPORT.md puts /research-lit in composed mode (see Output hygiene above): it returns the landscape for folding into the report instead of writing a standalone landscape file. The report doesn't exist yet at Phase 1 — the directive names the forthcoming canonical doc, and /idea-creator creates it in Phase 2.

If gemini-cli is not installed, /research-lit skips the Gemini source gracefully with a warning — no break to the pipeline. Users who want to force-disable Gemini in idea-discovery can pass /idea-discovery "topic" — sources: all explicitly (which becomes the literal source list, no auto-injection).

What this does:

• Search arXiv, Google Scholar, Semantic Scholar for recent papers
• Plus Gemini-driven broad discovery (sub-problem decomposition, naming variants, alias coverage) when gemini-cli is available
• Build a landscape map: sub-directions, approaches, open problems
• Identify structural gaps and recurring limitations
• Output a literature summary (saved to working notes)

🚦 Checkpoint: Present the landscape summary to the user.

When AUTO_PROCEED=true (non-blocking): report the selected direction and continue immediately in the same turn, without a question:

📚 Literature survey complete. Here's what I found:
- [key findings, gaps, open problems]

AUTO_PROCEED: selected [top-ranked direction]. Continuing to Phase 2.

When AUTO_PROCEED=false (blocking): present the same findings, ask Does this match your understanding? Should I adjust the scope before generating ideas?, then end the turn.

• User approves → proceed to Phase 2 with the best direction.
• User requests changes (e.g., "focus more on X", "ignore Y", "too broad") → refine the search with updated queries, re-run /research-lit with adjusted scope, and present again. Repeat until the user is satisfied.

Phase 2: Idea Generation + Filtering + Pilots

Invoke /idea-creator with the landscape context (and idea-stage/REF_PAPER_SUMMARY.md if available):

/idea-creator "$ARGUMENTS" — composed: idea-stage/IDEA_REPORT.md

/idea-creator owns idea-stage/IDEA_REPORT.md as the canonical deliverable; the — composed: directive tells it to fold the survey/novelty findings in rather than emitting LIT_LANDSCAPE.md / RESEARCH_REVIEW.md / MANIFEST.md alongside.

What this does:

• If idea-stage/REF_PAPER_SUMMARY.md exists, include it as context — ideas should build on, improve, or extend the reference paper
• Brainstorm 8-12 concrete ideas via GPT-5.6-Sol xhigh
• Filter by feasibility, compute cost, quick novelty search
• Deep validate top ideas (full novelty check + devil's advocate)
• Run parallel pilot experiments on available GPUs (top 2-3 ideas)
• Rank by empirical signal
• Output idea-stage/IDEA_REPORT.md

🚦 Checkpoint: Present idea-stage/IDEA_REPORT.md ranked ideas to the user.

When AUTO_PROCEED=true (non-blocking): report the automatic selection and continue immediately in the same turn, without a question:

💡 Generated X ideas, filtered to Y, piloted Z. Top results:

1. [Idea 1] — Pilot: POSITIVE (+X%)
2. [Idea 2] — Pilot: WEAK POSITIVE (+Y%)
3. [Idea 3] — Pilot: NEGATIVE, eliminated

AUTO_PROCEED: selected [top-ranked idea(s)]. Continuing to Phase 3.

When AUTO_PROCEED=false (blocking): present the same ranking, ask Which ideas should I validate further? Or should I regenerate with different constraints?, then end the turn.

• User picks ideas → proceed to Phase 3 with the selected ideas.
• User unhappy with all ideas → collect feedback ("what's missing?", "what direction do you prefer?"), update the prompt with user's constraints, and re-run Phase 2 (idea generation). Before regenerating, read the already-tried directions (research-wiki Failed Ideas + any .aris/runs/<run_id>.iterations.jsonl) and forbid a candidate too close to one already tried — enforced direction diversity; when an overnight heartbeat drives the run, record each chosen direction via iteration_log.py note ... --direction "<frame>" so later ticks can reject near-duplicates (see shared-references/external-cadence.md → Stall detection & forced structural pivot). Repeat until the user selects at least 1 idea.
• User wants to adjust scope → go back to Phase 1 with refined direction.

Phase 3: Deep Novelty Verification

For each top idea (positive pilot signal), run a thorough novelty check:

/novelty-check "[top idea 1 description]"
/novelty-check "[top idea 2 description]"

What this does:

• Multi-source literature search (arXiv, Scholar, Semantic Scholar)
• Cross-verify with GPT-5.6-Sol xhigh
• Check for concurrent work (last 3-6 months)
• Identify closest existing work and differentiation points

Update idea-stage/IDEA_REPORT.md with deep novelty results. Eliminate any idea that turns out to be already published.

Phase 4: External Critical Review

For the surviving top idea(s), get a sharp outside read — strongest case, named risks, and the cheapest discriminating next experiment; the core hypothesis is not up for rewriting:

/research-review "[top idea with hypothesis + pilot results]" — composed: idea-stage/IDEA_REPORT.md

In composed mode /research-review folds its conclusions into idea-stage/IDEA_REPORT.md and cites the .aris/traces/… path instead of writing a standalone review .md in the project root.

What this does:

• GPT-5.6-Sol xhigh acts as a senior reviewer (NeurIPS/ICML level)
• Scores the idea, identifies weaknesses, suggests minimum viable improvements
• Provides concrete feedback on experimental design

Update idea-stage/IDEA_REPORT.md with reviewer feedback and revised plan.

Phase 4.5: Method Refinement + Experiment Planning

After review, refine the top idea into a concrete proposal and plan experiments:

/research-refine-pipeline "[top idea description + pilot results + reviewer feedback]"

What this does:

• Freeze a Problem Anchor to prevent scope drift
• Refine the method via GPT-5.6-Sol review — reviewer risks choose the next tests, they do not add components; the score is advisory, and preserving the core hypothesis outranks pleasing the reviewer
• Generate a claim-driven experiment roadmap with ablations, budgets, and run order
• Output: refine-logs/FINAL_PROPOSAL.md, refine-logs/EXPERIMENT_PLAN.md, refine-logs/EXPERIMENT_TRACKER.md

🚦 Checkpoint: Present the refined proposal summary.

When AUTO_PROCEED=true (non-blocking): report that the proposal was selected and continue immediately in the same turn, without a question:

🔬 Method refined and experiment plan ready:
- Problem anchor: [anchored problem]
- Method thesis: [one sentence]
- Dominant contribution: [what's new]
- Must-run experiments: [N blocks]
- First 3 runs to launch: [list]

AUTO_PROCEED: accepted the top proposal. Continuing to Final Report.

When AUTO_PROCEED=false (blocking): present the same summary, ask Proceed to implementation? Or adjust the proposal?, then end the turn.

• User approves → proceed to Final Report.
• User requests changes → pass feedback to /research-refine for another round.
• Lite mode: If the pilot was inconclusive, still produce the smallest discriminating next-experiment plan — a reviewer score alone never downgrades an idea.

Phase 5: Final Report

Finalize idea-stage/IDEA_REPORT.md with all accumulated information:

# Idea Discovery Report

**Direction**: $ARGUMENTS
**Date**: [today]
**Pipeline**: research-lit → idea-creator → novelty-check → research-review → research-refine-pipeline

## Executive Summary
[2-3 sentences: best idea, key evidence, recommended next step]

## Literature Landscape
[from Phase 1]

## Ranked Ideas
[from Phase 2, updated with Phase 3-4 results]

## Novelty Verification
[from Phase 3]

## External Critical Review
[from Phase 4]

### 🏆 Idea 1: [title] — RECOMMENDED
- Pilot: POSITIVE (+X%)
- Novelty: CONFIRMED (closest: [paper], differentiation: [what's different])
- Reviewer score: X/10
- Next step: implement full experiment → /auto-review-loop

### Idea 2: [title] — BACKUP
...

## Eliminated Ideas
[ideas killed at each phase, with reasons]

## Refined Proposal
- Proposal: `refine-logs/FINAL_PROPOSAL.md`
- Experiment plan: `refine-logs/EXPERIMENT_PLAN.md`
- Tracker: `refine-logs/EXPERIMENT_TRACKER.md`

## Next Steps
- [ ] /run-experiment to deploy experiments from the plan
- [ ] /auto-review-loop to iterate until submission-ready
- [ ] Or invoke /research-pipeline for the complete end-to-end flow

Before presenting this report as complete, run the per-stage evidence gate above. A BLOCKED gate result is part of the report, not a warning to omit.

Phase 5.5: Write Compact Files (when COMPACT = true)

Skip entirely if COMPACT is false.

Write idea-stage/IDEA_CANDIDATES.md — a lean summary of the top 3-5 surviving ideas:

# Idea Candidates

| # | Idea | Pilot Signal | Novelty | Reviewer Score | Status |
|---|------|-------------|---------|---------------|--------|
| 1 | [title] | +X% | Confirmed | X/10 | RECOMMENDED |
| 2 | [title] | +Y% | Confirmed | X/10 | BACKUP |
| 3 | [title] | Negative | — | — | ELIMINATED |

## Active Idea: #1 — [title]
- Hypothesis: [one sentence]
- Key evidence: [pilot result]
- Next step: /experiment-bridge or /research-refine

This file is intentionally small (~30 lines) so downstream skills and session recovery can read it without loading the full idea-stage/IDEA_REPORT.md (~200+ lines).

Phase 5.6: Instantiate the Research Contract (always — NOT gated on COMPACT)

When Phase 4 ends with a RECOMMENDED idea, create idea-stage/docs/research_contract.md from templates/RESEARCH_CONTRACT_TEMPLATE.md (resolve the template from the repo root or $ARIS_REPO/templates/), filling in: the selected idea + selection rationale, core claims, minimum convincing evidence, and the next-step pointer. Skip only when the run produced no RECOMMENDED idea.

This file is the focused working contract for the W1 → W1.5 handoff: /experiment-bridge implements against it, and /result-to-claim + /ablation-planner read it as the claims source. It is also the #2 session-recovery file (docs/SESSION_RECOVERY_GUIDE.md) — a crashed session reloads the ACTIVE idea from this contract instead of the full idea pool.

Output Protocols

Follow these shared protocols for all output files:

• Output Composition Protocol — ONE canonical deliverable per pipeline; fold sub-skill findings in, don't scatter overlapping .md files
• Output Versioning Protocol — write timestamped file first, then copy to fixed name
• Output Manifest Protocol — maintain MANIFEST.md only above the 15-artifact threshold (not "log every output")
• Output Language Protocol — respect the project's language setting

Output hygiene — ONE canonical doc, no duplicate MDs (REQUIRED)

This pipeline runs its sub-skills in composed mode (see output-composition.md): it owns a single canonical deliverable and folds every sub-skill's findings into it rather than letting each emit its own overlapping file. Concretely, for this workflow:

1. idea-stage/IDEA_REPORT.md is the single canonical deliverable. Sub-skills' intermediate findings (literature landscape, novelty notes, external review) are folded into it as sections/appendices — they do NOT become standalone files just because a sub-skill could emit one. If a sub-skill writes a scratch file, inline its unique content into the report and delete the scratch when the phase closes.
2. Pass — composed: idea-stage/IDEA_REPORT.md to every sub-skill (/research-lit, /idea-creator, /research-review) so they fold instead of scatter. This is the explicit signal; without it a sub-skill stays standalone by design.
3. Refined-method outputs stay in refine-logs/ (FINAL_PROPOSAL.md / EXPERIMENT_PLAN.md / EXPERIMENT_TRACKER.md). Do NOT also restate them as separate files under idea-stage/; the report links to them, it does not copy them.
4. No MANIFEST.md for a handful of files — only above the 15-artifact threshold in output-manifest.md.
5. Pilot scratch is disposable: keep the pilot script (reusable) + one results file (pilot_results.jsonl or a small summary). Delete launcher logs, smoke files, and redundant *_summary.json once the numbers are in the report.
6. Cross-model review traces belong in .aris/traces/… (the audit trail); do not ALSO keep a human-facing copy under idea-stage/ — cite the trace path from the report.
7. Before finishing, the idea-stage/ top level should be roughly: IDEA_REPORT.md (+ .html), the pilot script + results, and the refine-logs/ dir. Nothing else unless it carries content not in the report.

Render HTML view (auto, when RENDER_HTML = true)

After Phase 4 finalizes idea-stage/IDEA_REPORT.md (and the optional IDEA_CANDIDATES.md), invoke /render-html on the report so the user has a single-file HTML view for tablet / phone reading:

/render-html "idea-stage/IDEA_REPORT.md" --no-review

--no-review is intentional: source MD already passed this skill's own novelty + cross-model review. HTML render is a structural conversion, not a new claim-audit gate. Output lands at idea-stage/IDEA_REPORT.html with embedded source SHA256 + render timestamp.

Non-blocking: if /render-html fails (helper missing, Codex MCP unavailable, file write error), log the failure and continue — the HTML view is a convenience artifact, not a Phase 4 prerequisite.

Skip this step if RENDER_HTML = false.

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.

• Don't skip phases. Each phase filters and validates — skipping leads to wasted effort later.

• Checkpoint between phases. Briefly summarize what was found. With AUTO_PROCEED=true, state the selected next action and keep executing in the same turn; with false, ask and end the turn.

• Let pilots kill, not vibes. A cheap pilot that says no beats a month of implementation that says no — but the kill needs empirical signal or a named published paper, not taste. Talking yourself out of ideas on paper is how pipelines end up with nothing to run.

• Empirical signal > theoretical appeal. An idea with a positive pilot outranks a "sounds great" idea without evidence.

• Document everything — inside the one report, not in scattered files. Dead ends and eliminated ideas are valuable, so record them as sections of idea-stage/IDEA_REPORT.md (see Output hygiene above). Do not spawn a separate .md per phase.

• Be honest with the reviewer. Include negative results and failed pilots in the review prompt.

• Feishu notifications are optional. If ~/.claude/feishu.json exists, send checkpoint at each phase transition and pipeline_done at final report. If absent/off, skip silently.

Composing with Workflow 2

After this pipeline produces a validated top idea:

/idea-discovery "direction"         ← you are here (Workflow 1, includes method refinement + experiment planning)
/run-experiment                     ← deploy experiments from the plan
/auto-review-loop "top idea"        ← Workflow 2: iterate until submission-ready

Or use /research-pipeline for the full end-to-end flow.