Skill

review-pr-prose

From idh

Simulated peer review panel for manuscript prose. Spins discipline-specific agents for multi-perspective review.

Install

npx claudepluginhub minhhaduong/imperialdragonharness

Tool Access

This skill uses the workspace's default tool permissions.

Preview

Spin disciplinary agents in parallel, each in a fresh context. Prose review reads **full text**, not just diff.

SKILL.md

Similar Skills

kotlin-ktor-patterns

Provides Ktor server patterns for routing DSL, plugins (auth, CORS, serialization), Koin DI, WebSockets, services, and testApplication testing.

everything-claude-code

163.2k

deep-research

Conducts multi-source web research with firecrawl and exa MCPs: searches, scrapes pages, synthesizes cited reports. For deep dives, competitive analysis, tech evaluations, or due diligence.

everything-claude-code

163.2k

inventory-demand-planning

Provides demand forecasting, safety stock optimization, replenishment planning, and promotional lift estimation for multi-location retailers managing 300-800 SKUs.

everything-claude-code

163.2k

Stats

Stars0

Forks0

Last CommitApr 15, 2026

Actions

View Source View Plugin View on GitHub View README

Review PR prose $ARGUMENTS — simulated peer review panel

Spin disciplinary agents in parallel, each in a fresh context. Prose review reads full text, not just diff.

Setup

Identify the text: which .qmd/.md files changed? What is the target venue?

Read the diff of the merge request.

Recruit the panel: select agents appropriate for the venue and scope of changes. Always include an adversarial referee. Add a journal-specific expert if venue rules exist (check project rules).

Each agent runs

Read the full text (not just the diff).

Report confidence + severity (major / minor / suggestion).

Verdict: accept, minor revision, or major revision.

Agents with relevant expertise should use available tools (web search for literature, linting tools if installed, etc.).

AI-tells auditor (always included)

One agent is always the AI-tells auditor. It reads config/ai-tells.yml for blacklisted words, phrases, conditional words, density limits, and patterns to flag. It scans the full text (not just the diff) and reports every violation with line number, context, and severity. This agent has no other role — it is a specialized lint pass.

Synthesis

Preserve dissent verbatim.

Group findings: major (blocks acceptance), minor (should fix), suggestion.

Deduplicate convergent findings.

Build the manuscript. Check consistency between prose and data.

Post a single review on the merge request.

Proportional depth

Text change

Panel size

Typo, citation fix

Copy editor only

Section rewrite

3 agents (domain + adversarial + copy)

Full paper draft

Full panel (5-6 agents)

Submission-ready

Full panel + response-to-reviewers template

Review PR prose $ARGUMENTS — simulated peer review panel

Spin disciplinary agents in parallel, each in a fresh context. Prose review reads full text, not just diff.

Setup

Identify the text: which .qmd/.md files changed? What is the target venue?
Read the diff of the merge request.
Recruit the panel: select agents appropriate for the venue and scope of changes. Always include an adversarial referee. Add a journal-specific expert if venue rules exist (check project rules).

Each agent runs

Read the full text (not just the diff).
Report confidence + severity (major / minor / suggestion).
Verdict: accept, minor revision, or major revision.

Agents with relevant expertise should use available tools (web search for literature, linting tools if installed, etc.).

AI-tells auditor (always included)

Synthesis

Preserve dissent verbatim.
Group findings: major (blocks acceptance), minor (should fix), suggestion.
Deduplicate convergent findings.
Build the manuscript. Check consistency between prose and data.
Post a single review on the merge request.

Proportional depth

Text change	Panel size
Typo, citation fix	Copy editor only
Section rewrite	3 agents (domain + adversarial + copy)
Full paper draft	Full panel (5-6 agents)
Submission-ready	Full panel + response-to-reviewers template