Ready to supercharge your AI agents?

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

Respond to Referee · brycewang-stanford | DecimalAI

researchby brycewang-stanford

Respond to Referee

brycewang-stanford/respond-to-referee

Source brycewang-stanford/Auto-Empirical-Research-Skills ↗License NOASSERTION

Structure point-by-point referee responses for the sewage-house-prices paper. Classifies each comment (NEW ANALYSIS / CLARIFICATION / REWRITE / DISAGREE / MINOR), produces a tracking document, drafts a response letter in LaTeX, and flags items needing new analysis or user judgment. This skill should be used when asked to "respond to referees", "draft revision", "address referee comments", or "R&R".

SkillScore88Maintained

Automatic

Research pack

Last verified 8/8/2026

What it changes

Passed 18 of 20 test cases with the skill, and 12 of 20 without it.

Without skill58%

With skill88%

Checks it passed that the model alone missed

The response references section files in docs/overleaf/.
The response includes total count of required new analyses in the summary metrics.
The response identifies docs/overleaf/_main.tex as the primary manuscript LaTeX document.

and 5 more in the full report

7 fixed14 passed either way3 still failing0 broke

Pass-rate lift: +29 pp
Tokens: +23%
Agent turns: 0%
Response time: -37%

Model gemini-3.6-flashMeasured 8/8/2026v121 of 24 cases passed with skill · 20 comparable

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/brycewang-stanford-respond-to-referee/SKILL.md https://app.decimal.ai/s/brycewang-stanford-respond-to-referee/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add brycewang-stanford-respond-to-referee, 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/brycewang-stanford-respond-to-referee/ and pins it in .decimal/skills.lock

decimalai skills add brycewang-stanford-respond-to-referee

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

curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-respond-to-referee/SKILL.md https://app.decimal.ai/s/brycewang-stanford-respond-to-referee/SKILL.md

Raw file: /s/brycewang-stanford-respond-to-referee/SKILL.md ↗ · machine-readable summary at /s/brycewang-stanford-respond-to-referee.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 6

Prompt

When checking existing econometric scripts to verify if a reviewer's suggested specification has already been run, where are the R analysis scripts located?

What was checked, without and with the skill
Checked	Without skill	With skill
The response identifies scripts/R/09_analysis/ as the directory containing existing econometric analysis scripts.	Missed	Passed

case-11 · gemini-3.6-flash · 8/8/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md154 linessingle file

Respond to Referee

Structure a point-by-point response to referee reports for the sewage-house-prices manuscript.

Input: $ARGUMENTS — path to referee report file(s).

Project-Specific Context

Manuscript

Main document: docs/overleaf/_main.tex
Sections: docs/overleaf/*.tex
Bibliography: docs/overleaf/refs.bib

Analysis

Scripts: scripts/R/09_analysis/
Tables: output/tables/
Figures: output/figures/

Common Referee Concerns for This Paper

Short EDM panel (2021-2024) — limited time variation
LSOA FE may not capture neighbourhood-level sorting
Spill count vs hours as treatment — which matters more?
External validity beyond England
Dry spill identification — rainfall threshold sensitivity
Multiple testing across 6 radii and multiple specifications

Workflow

Step 1: Parse Inputs

Read referee report(s) from $ARGUMENTS
Read the manuscript (docs/overleaf/_main.tex and section files)
Scan existing analysis scripts to know what already exists

Step 2: Classify Every Comment

Track record

Benchmark

18 / 20

test cases passed, vs 12 / 20 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 — 8/2/2026

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

brycewang-stanford / Auto-Empirical-Research-Skills

skills/41-sticerd-eee-sewage-econometrics-check/skills/respond-to-referee/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/2/2026Version v1

Other skills for the same job

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

brycewang-stanford/reviseR&R cycle — classify referee comments and route to appropriate agents. Replaces /respond-to-referee.Close match+27 pp pass-rate 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

Respond to Referee

Project-Specific Context

Manuscript

Analysis

Common Referee Concerns for This Paper

Workflow

Step 1: Parse Inputs

Step 2: Classify Every Comment

Track record

Benchmark

Adoption

SkillSafety

Source

Other skills for the same job

Related skills