novelty-check

name	novelty-check
version	0.1.0
description	Compare your research idea against existing literature, assess novelty, and refine the hypothesis. Produces refined-idea.md that downstream skills consume. Use when asked to "check novelty", "is this novel", "compare against existing work", or "refine my idea". Proactively invoke after /lit-review completes.
allowed-tools	["Bash","Read","Write","Edit","Grep","Glob","WebSearch","AskUserQuestion"]

Preamble (run first)

# --- GENERATED PREAMBLE START ---
_PROJECT_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || pwd)
if ! git rev-parse --show-toplevel >/dev/null 2>&1; then
  echo "WARNING: Not inside a git repository. Files will be written to $(pwd)."
fi
mkdir -p ~/.rstack/sessions ~/.rstack/analytics "$_PROJECT_ROOT/.rstack"
touch ~/.rstack/sessions/"$PPID"
find ~/.rstack/sessions -mmin +120 -type f -delete 2>/dev/null || true
_SESSIONS=$(find ~/.rstack/sessions -mmin -120 -type f 2>/dev/null | wc -l | tr -d ' ')
_RSTACK_DIR="$(cd "$(dirname "$0")/.." 2>/dev/null && pwd || echo "$HOME/.claude/skills/rstack")"
_RSTACK_CONFIG="$_RSTACK_DIR/bin/rstack-config"
_UPD=$("$_RSTACK_DIR/bin/rstack-update-check" 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
_VENUE=$("$_RSTACK_CONFIG" get venue 2>/dev/null || echo "arxiv")
_COMPUTE=$("$_RSTACK_CONFIG" get compute_preferred 2>/dev/null || echo "modal")
_PROACTIVE=$("$_RSTACK_CONFIG" get proactive 2>/dev/null || echo "true")
_TEL=$("$_RSTACK_CONFIG" get telemetry 2>/dev/null || echo "off")
_TEL_PROMPTED=$([ -f ~/.rstack/.telemetry-prompted ] && echo "yes" || echo "no")
_BRANCH=$(git branch --show-current 2>/dev/null || echo "unknown")
echo "PROJECT_ROOT: $_PROJECT_ROOT"
echo "BRANCH: $_BRANCH"
echo "VENUE: $_VENUE"
echo "COMPUTE: $_COMPUTE"
echo "PROACTIVE: $_PROACTIVE"
if [ ! -f ~/.rstack/.setup-complete ]; then
  echo "NEEDS_SETUP"
fi
_TEL_START=$(date +%s)
_SESSION_ID="$$-$(date +%s)"
echo "TELEMETRY: ${_TEL:-off}"
echo "TEL_PROMPTED: $_TEL_PROMPTED"
if [ "$_TEL" != "off" ]; then
  echo '{"skill":"novelty-check","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'","repo":"'$(basename "$(git rev-parse --show-toplevel 2>/dev/null)" 2>/dev/null || echo "unknown")'"}'  >> ~/.rstack/analytics/skill-usage.jsonl 2>/dev/null || true
fi
if [ "$_TEL" != "off" ]; then
  echo '{"skill":"novelty-check","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'","session_id":"'"$_SESSION_ID"'","rstack_version":"'"$(cat "$_RSTACK_DIR/VERSION" 2>/dev/null | tr -d "[:space:]" || echo "unknown")"'"}'  > ~/.rstack/analytics/.pending-"$_SESSION_ID" 2>/dev/null || true
fi
for _PF in $(find ~/.rstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do
  if [ -f "$_PF" ]; then
    _PF_BASE="$(basename "$_PF")"
    _PF_SID="${_PF_BASE#.pending-}"
    [ "$_PF_SID" = "$_SESSION_ID" ] && continue
    if [ "$_TEL" != "off" ] && [ -x "$_RSTACK_DIR/bin/rstack-telemetry-log" ]; then
      "$_RSTACK_DIR/bin/rstack-telemetry-log" --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true
    fi
    rm -f "$_PF" 2>/dev/null || true
  fi
  break
done
eval "$("$_RSTACK_DIR/bin/rstack-slug" 2>/dev/null)" 2>/dev/null || true
echo "SLUG: ${SLUG:-unknown}"
# --- GENERATED PREAMBLE END ---

If output shows UPGRADE_AVAILABLE <old> <new>: read rstack-upgrade/SKILL.md and follow the "Inline Upgrade Flow". Then continue with this skill. If output shows JUST_UPGRADED <from> <to>: tell user "Running RStack v{to} (just updated!)" and continue.

If output shows NEEDS_SETUP: tell user "RStack not configured yet. Run /setup to configure compute providers." Then read setup-skill/SKILL.md and follow it inline before continuing.

If TEL_PROMPTED is no: Ask the user about telemetry. Use AskUserQuestion:

Help RStack get better! Community mode shares usage data (which skills you use, how long they take, crash info) with a stable device ID so we can track trends and fix bugs faster. No code, file paths, or repo names are ever sent. Change anytime with rstack-config set telemetry off.

Options:

• A) Help RStack get better! (recommended)
• B) No thanks

If A: run rstack-config set telemetry community

If B: ask a follow-up:

How about anonymous mode? Just a counter that helps us know if anyone's out there.

Options:

• A) Sure, anonymous is fine
• B) No thanks, fully off

If B→A: run rstack-config set telemetry anonymous If B→B: run rstack-config set telemetry off

Always run:

touch ~/.rstack/.telemetry-prompted

This only happens once. If TEL_PROMPTED is yes, skip this entirely.

Important: Note the PROJECT_ROOT value from the preamble output. All file paths below are relative to this project root directory. Work products (idea.md, refined-idea.md, etc.) go at the project root. Plumbing (.rstack/lit-review.jsonl) goes in the .rstack/ subdirectory.

When to Use

Run this after /lit-review to assess whether your research idea is novel. Can also run standalone — if no literature review exists, it runs a lightweight search first (top 10 papers only).

Step 0: Load Context

1. Read idea.md at the project root. If it does not exist, ask the user for their research idea and save it.
2. Check if .rstack/lit-review.jsonl exists.
   • If YES: read it. This is the primary source for comparison.
   • If NO: run a lightweight search. Use WebSearch to find the top 10 most relevant papers for this idea. Record them as temporary context. Note: "Running lightweight novelty check without full literature review. Run /lit-review first for higher confidence."

Step 1: Paper-by-Paper Comparison

For each highly relevant paper (relevance >= 7 from lit-review.jsonl, or all papers from lightweight search):

Compare the research idea against the paper on these dimensions:

• Method overlap: Does the paper use the same approach? Same model architecture, training method, loss function?
• Application overlap: Does the paper target the same domain/task?
• Dataset overlap: Does the paper use the same benchmarks?
• Contribution overlap: Does the paper make the same claim about what works and why?

For each paper, classify the relationship:

• direct_competitor — same method, same domain
• partial_overlap — shares method OR domain but not both
• complementary — different approach to the same problem (potential baseline)
• foundational — builds the theory our work extends
• unrelated — low relevance, skip

Step 2: Novelty Assessment

Synthesize the paper comparisons into a novelty assessment:

1. What is genuinely novel about this idea? (Be specific — not "we apply X to Y" but "we combine X's loss function with Y's architecture, which no existing work has tried because...")
2. What overlaps with existing work? (Name specific papers)
3. What gaps in existing work does this idea address?
4. Novelty score: 1-10 where:
   • 1-3: Significant overlap with existing work, marginal contribution
   • 4-6: Incremental improvement over existing methods
   • 7-8: Genuine novelty in method, application, or insight
   • 9-10: Paradigm-shifting new approach

Write assessment to novelty-assessment.md at the project root.

Step 3: Idea Refinement

Based on the novelty assessment, produce a refined research statement. This is critical — /experiment consumes refined-idea.md, not the raw idea.md.

Write refined-idea.md at the project root with these sections (extracted from Ignis refine-idea-agent.ts):

# Refined Research Idea

## Problem Statement
[One paragraph: what problem are we solving and why it matters]

## Proposed Approach
[2-3 paragraphs: what we will do, technically specific]

## What is Novel
[Bullet list: exactly what contributions are new vs existing work]

## Key Assumptions
[What must be true for this approach to work]

## Evaluation Plan
[What metrics, what datasets, what baselines to compare against]

## Risks
[What could go wrong, what might not work]

## Next Actions
[Concrete next steps for /experiment]

Step 4: Human Checkpoint

Use AskUserQuestion:

Project: {project name}, branch: {branch}

Novelty assessment complete. Your idea scores {score}/10 for novelty. {one sentence summary of the main novel contribution}. {one sentence about the main overlap concern, if any}.

The refined hypothesis is saved to refined-idea.md.

RECOMMENDATION: Choose A to proceed with experiments.

Options:

• A) Proceed to experiments — the idea is novel enough
• B) Refine further — I want to adjust the approach (loops back to Step 3)
• C) Deepen literature — search for more papers in area X (loops back to /lit-review)
• D) Pivot — the idea needs fundamental rethinking (loops back to idea.md)

If B: ask what to change, update refined-idea.md, re-assess novelty. If C: suggest running /lit-review with more specific queries. If D: ask for the new direction, update idea.md, re-run from Step 0.

Important Rules

• Never tell the user their idea is novel when it clearly overlaps with existing work. Be honest.
• Always name specific papers when discussing overlap. "Your approach resembles Smith et al. (2024)" not "there is some overlap."
• If novelty score is below 4, strongly recommend pivoting or finding a sharper angle before spending GPU hours.
• refined-idea.md is the most important output. It must be specific enough that /experiment can generate code from it.

---

Telemetry (run last)

After the skill workflow completes (success, error, or abort), log the telemetry event.

# --- GENERATED EPILOGUE START ---
_TEL_END=$(date +%s)
_TEL_DUR=$(( _TEL_END - _TEL_START ))
rm -f ~/.rstack/analytics/.pending-"$_SESSION_ID" 2>/dev/null || true
if [ "$_TEL" != "off" ]; then
  echo '{"skill":"novelty-check","duration_s":"'"$_TEL_DUR"'","outcome":"OUTCOME","session":"'"$_SESSION_ID"'","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'"}' >> ~/.rstack/analytics/skill-usage.jsonl 2>/dev/null || true
fi
if [ "$_TEL" != "off" ] && [ -x "$_RSTACK_DIR/bin/rstack-telemetry-log" ]; then
  "$_RSTACK_DIR/bin/rstack-telemetry-log" \
    --skill "novelty-check" --duration "$_TEL_DUR" --outcome "OUTCOME" \
    --session-id "$_SESSION_ID" --pipeline-stage "novelty-check" 2>/dev/null &
fi
# --- GENERATED EPILOGUE END ---

Replace OUTCOME with success/error/abort based on the workflow result.