domains / research / external-research-review / research-review

research-review

Skill card

combined

The two copies implement the same multi-round external research review workflow, but they disagree on the reviewer backend, model, reasoning effort, MCP setup, and prompt structure. The wanshuiyin copy is a more evolved, backend-routed version (Codex or Manual) with a file-based brief, adversarial framing, and explicit scope limits. The brycewang copy is a simpler Codex-only version with inline context, xhigh reasoning, and a set of follow-up prompt templates. Neither is a strict reproduction of the other; crowning one would hide a real alternative (manual backend vs Codex-only, file brief vs inline context, ultra vs xhigh).

Repositories

Other copies

Combined skill

---
name: research-review
description: Get a deep critical review of research from an external reviewer backend (Codex or manual). Use when user says "review my research", "help me review", "get external review", or wants critical feedback on research ideas, papers, or experimental results.
argument-hint: "[topic-or-scope]"
allowed-tools: Bash(*), Read, Grep, Glob, Write, Edit, mcp__codex__codex, mcp__codex__codex-reply, mcp__manual_review__review, mcp__manual_review__review_reply
---

# Research Review via External Reviewer Backend

> 🔒 **Do not wrap this skill in `/loop`, `/schedule`, or `CronCreate`.** It is verdict-bearing — it produces a cross-model review verdict, multi-round with reviewer thread continuity. An external timer re-fires the verdict on wall-clock time and breaks the reviewer's round-to-round memory: zero new signal, full token cost. Schedule the *external wait that precedes it* (work ready → then review once), not the verdict.

Get a multi-round critical review of research work from the selected external reviewer backend with maximum reasoning depth.

## Constants

- REVIEWER_MODEL = `gpt-6-astra` — Default model for the Codex backend, reasoning effort `ultra` (deep-audit tier). Must be an OpenAI model (e.g., `gpt-6-astra`, `gpt-5.5`, `o3`). Manual backend uses a model the user chooses — it must be a recognized model from a different family (OpenAI, Anthropic, Google, DeepSeek, Moonshot/Kimi, Qwen).
- **REVIEWER_BACKEND = `codex`** — Default: Codex MCP (ultra). Override with `— reviewer: oracle-pro` for Oracle MCP, or `— reviewer: manual` for Manual Review MCP. If manual-review MCP is unavailable, stop and print the install command; do not fall back to Codex.

## Reviewer Calling Convention

When calling the reviewer, branch on REVIEWER_BACKEND:

**If REVIEWER_BACKEND = `codex`:**
  Use `mcp__codex__codex` for new review threads.
  Use `mcp__codex__codex-reply` for follow-up rounds (reuse threadId).

**If REVIEWER_BACKEND = `manual`:**
  Use `mcp__manual_review__review` for new review threads with:
    prompt: [exact same prompt that would go to Codex]
    config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}
  Save the returned `threadId`.
  Use `mcp__manual_review__review_reply` for follow-up rounds with:
    threadId: [saved manual-review threadId]
    prompt: [follow-up prompt]
    config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}

Content fidelity: the manual reviewer should see the same substantive review brief Codex would read. If the manual UI supports file upload / attachment, reuse the same brief file; otherwise paste the brief contents inline because remote web UIs cannot read your local filesystem paths.

## Context: $ARGUMENTS

## Prerequisites

- **Codex MCP Server** configured in Claude Code:
  ```bash
  claude mcp add codex -s user -- python3 "$HOME/aris_repo/mcp-servers/codex-exec/server.py"   # your ARIS clone's path
  ```
- This gives Claude Code access to `mcp__codex__codex` and `mcp__codex__codex-reply` tools

## Workflow

### Step 1: Gather Research Context
Before calling the external reviewer, compile a comprehensive briefing:
1. Read project narrative documents (e.g., STORY.md, README.md, paper drafts)
2. Read any memory/notes files for key findings and experiment history
3. Identify: core claims, methodology, key results, known weaknesses

### Step 2: Initial Review (Round 1)
Send a detailed prompt with ultra reasoning, using the selected backend. For the `codex` backend, keep the MCP payload short: write the full briefing to `RESEARCH_REVIEW_REQUEST.md`, then point Codex at that file.

*For codex backend:*

```
mcp__codex__codex:
  model: gpt-6-astra
  config: {"model_reasoning_effort": "ultra"}
  prompt: |
    Read the review brief at <absolute path to RESEARCH_REVIEW_REQUEST.md>.
    Executor notes are not evidence beyond the files they cite, so verify the
    referenced artifacts before judging.
    Please act as a senior ML reviewer (NeurIPS/ICML level). Start from the
    assumption that the work is broken somewhere — your job is to find where.
    Be adversarial. Trust nothing the author tells you — verify everything
    yourself. Identify:
    1. Logical gaps or unjustified claims
    2. Missing experiments that would strengthen the story
    3. Narrative weaknesses
    4. Whether the contribution is sufficient for a top venue

    === SCOPE LIMITS (these bound what you PROPOSE, never what you look for) ===
    Report anything that is actually wrong here — including a rare-looking case, if
    this repo actually produces it. Then keep the fix in scope:
    1. This is a RESEARCH-WORKFLOW tool, not a security paper. Verification is
       welcome; over-defense is not. Assume a cooperating operator on their own
       machine — a malicious local user is NOT in the threat model.
    2. Do NOT propose SHA / hash / content-fingerprint / digest-binding schemes.
       Reporting a real defect in hashing code that already exists is fine.
    3. NO speculative machinery: do not add feature flags, migration frameworks,
       compat layers, wrappers, pins, or similar mechanisms unless evidence shows
       a current repo defect they fix or an explicit existing invariant they must
       preserve. "Load-bearing", "compatibility", and "not scaffolding" are labels,
       not evidence. Point to the failing path/artifact or invariant, and check the
       proposal's factual premises, such as whether a named package version exists.
    4. NO corner-case obsession: exotic encodings, symlink races, RTL text and
       millisecond races are out of scope unless you can show the case arises here.
    5. Where a fix is proposed, it must be minimal and directly tied to a
       demonstrated defect or invariant.
```

*For manual backend:*

```
mcp__manual_review__review:
  prompt: [same prompt as above, with the brief file attached or pasted inline]
  config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}
```

### Step 3: Iterative Dialogue (Rounds 2-N)
Use `mcp__codex__codex-reply` (Codex) or `mcp__manual_review__review_reply` (manual) with the returned `threadId` to continue the conversation:

For each round:
1. **Respond** to criticisms with evidence/counterarguments
2. **Ask targeted follow-ups** on the most actionable points
3. **Request specific deliverables**: experiment designs, paper outlines, claims matrices

Key follow-up patterns:
- "If we reframe X as Y, does that change your assessment?"
- "What's the minimum experiment to satisfy concern Z?"
- "Please design the minimal additional experiment package (highest acceptance lift per GPU week)"
- "Please write a mock NeurIPS/ICML review with scores"
- "Give me a results-to-claims matrix for possible experimental outcomes"

### Step 4: Convergence
Stop iterating when:
- Both sides agree on the core claims and their evidence requirements
- A concrete experiment plan is established
- The narrative structure is settled

### Step 5: Document Everything
Save the full interaction and conclusions to a review document in the project root:
- Round-by-round summary of criticisms and responses
- Final consensus on claims, narrative, and experiments
- Claims matrix (what claims are allowed under each possible outcome)
- Prioritized TODO list with estimated compute costs
- Paper outline if discussed

Update project memory/notes with key review conclusions.

## Key Rules

- ALWAYS use the configured reasoning effort for reviews (ultra for Codex, xhigh for manual)
- Send comprehensive context in Round 1 — the external model cannot read your files (use file attachment for manual backend)
- Be honest about weaknesses — hiding them leads to worse feedback
- Push back on criticisms you disagree with, but accept valid ones
- Focus on ACTIONABLE feedback — "what experiment would fix this?"
- Document the threadId for potential future resumption
- The review document should be self-contained (readable without the conversation)

## Prompt Templates

### For initial review:
"I'm going to present a complete ML research project for your critical review. Please act as a senior ML reviewer (NeurIPS/ICML level)..."

### For experiment design:
"Please design the minimal additional experiment package that gives the highest acceptance lift per GPU week. Our compute: [describe]. Be very specific about configurations."

### For paper structure:
"Please turn this into a concrete paper outline with section-by-section claims and figure plan."

### For claims matrix:
"Please give me a results-to-claims matrix: what claim is allowed under each possible outcome of experiments X and Y?"

### For mock review:
"Please write a mock NeurIPS review with: Summary, Strengths, Weaknesses, Questions for Authors, Score, Confidence, and What Would Move Toward Accept."