Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Stage 1 · Breakdown Base Template · runtsang | DecimalAI

researchby runtsang

Stage 1 · Breakdown Base Template

runtsang/stage1-breakdown-template

Source runtsang/rebuttalstudio ↗License Not recorded

Base template for Stage1 reviewer-comment breakdown.

SkillScore87Maintained

Automatic

Research pack

Last verified 8/28/2026

What it changes

Passed 20 of 23 test cases with the skill, and 9 of 23 without it.

Without skill39%

With skill87%

Checks it passed that the model alone missed

Each weakness and question issue maps to a separate Response block.
The main top-level heading uses the exact pattern Stage1 ICLR Breakdown.
Weakness items use source_id values formatted as weakness1 and weakness2.

and 16 more in the full report

12 fixed8 passed either way2 still failing1 broke

Pass-rate lift: +48 pp
Tokens: +28%
Agent turns: 0%
Response time: -20%

Model gemini-3.6-flashMeasured 8/28/2026v120 of 23 cases passed with skill

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/runtsang-stage1-breakdown-template/SKILL.md https://app.decimal.ai/s/runtsang-stage1-breakdown-template/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add runtsang-stage1-breakdown-template, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/runtsang-stage1-breakdown-template/ and pins it in .decimal/skills.lock

decimalai skills add runtsang-stage1-breakdown-template

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/runtsang-stage1-breakdown-template/SKILL.md https://app.decimal.ai/s/runtsang-stage1-breakdown-template/SKILL.md

Raw file: /s/runtsang-stage1-breakdown-template/SKILL.md ↗ · machine-readable summary at /s/runtsang-stage1-breakdown-template.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 12 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 12

Prompt

Extract the summary and strengths sections from this ACL paper review: Summary: This paper introduces an innovative prompt-tuning method for multilingual translation. It achieves state-of-the-art results on 10 language pairs. Strengths: 1. Strong performance. 2. Thorough ablation analysis. Convert these into structured preserved sections, rephrasing the sentences to make them more professional and concise.

What was checked, without and with the skill
Checked	Without skill	With skill
The text of summary and strengths is copied verbatim without rephrasing or polished wording.	Missed	Passed

case-07 · gemini-3.6-flash · 8/28/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md99 linessingle file

Stage 1 · Breakdown Base Template

Use this template as the shared core for all Stage1 conference variants. Conference-specific files must apply this template first, then add only necessary overrides.

Shared input assumptions

Input is one reviewer's full text (OpenReview export, PDF copy, DOC/DOCX paste, etc.).
Review sections may appear as headings, bullets, numbered lists, or mixed prose.
A single bullet can contain multiple independent concerns.

Shared mandatory workflow

Extract conference-defined scores (numbers only)
- Extract every score required by the conference extension.
- Keep values numeric only (for example 6, 4, 3), without scale explanations.
- If a score is missing, leave it empty rather than inventing.
Preserve designated sections verbatim
- Keep conference-defined preserved sections unchanged.
- Keep line breaks where possible.
Split issue-bearing sections into atomic issues
- Split conference-defined weakness/question sources into issue-level items.
- One atomic issue should correspond to one independent concern or request.
- Split further if one item contains multiple asks.
Create response blocks (Response1 ... ResponseN)
- Each response block must include:
  - title: generated short subtitle
  - source: weakness or question
  - source_id: weaknessN or questionN
  - quoted_issue: exact verbatim text of the mapped atomic issue
Apply prefix and numbering rules
- Issues from weakness-type source sections must use weaknessN.
- Issues from question-type source sections must use questionN.
- Number weaknessN and questionN independently.

Shared splitting heuristics

Track record

Benchmark

20 / 23

test cases passed, vs 9 / 23 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

runtsang / rebuttalstudio

skills/stage1/template/SKILL.md

View on GitHub ↗

Imported 8/3/2026Version v1

Other skills for the same job

Published skills whose descriptions closely match this one, with their headline benchmark lift.

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Stage 1 · Breakdown Base Template

Shared input assumptions

Shared mandatory workflow

Shared splitting heuristics

Track record

Benchmark

Adoption

SkillSafety

Source

Other skills for the same job

Related skills