Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Tencent SkillHub Β· AI
Cross-Model Review
Adversarial plan review using two different AI models. Planner writes, reviewer challenges, they iterate until approved. Use when building features touching...
skill openclawclawhub Free
Adversarial plan review using two different AI models. Planner writes, reviewer challenges, they iterate until approved. Use when building features touching...
β¬ 0 downloads β
0 stars Unverified but indexed
Install for OpenClaw
Known item issue.
This item's current download entry is known to bounce back to a listing or homepage instead of returning a package file.
Quick setup
- Open the source page and confirm the package flow manually.
- Review SKILL.md if you can obtain the files.
- Treat this source as manual setup until the download is verified.
Package facts
- Download mode
- Manual review
- Package format
- ZIP package
- Source platform
- Tencent SkillHub
- What's included
- CHANGELOG.md, CONTRIBUTING.md, README.md, SECURITY.md, SKILL.md, package.json
Validation
- Open the source listing and confirm there is a real package or setup artifact available.
- Review SKILL.md before asking your agent to continue.
- Treat this source as manual setup until the upstream download flow is fixed.
Install with your agent
Agent handoff
Use the source page and any available docs to guide the install because the item currently does not return a direct package file.
- Open the source page via Open source listing.
- If you can obtain the package, extract it into a folder your agent can access.
- Paste one of the prompts below and point your agent at the source page and extracted files.
New install
I tried to install a skill package from Yavira, but the item currently does not return a direct package file. Inspect the source page and any extracted docs, then tell me what you can confirm and any manual steps still required. Then review README.md for any prerequisites, environment setup, or post-install checks.
Upgrade existing
I tried to upgrade a skill package from Yavira, but the item currently does not return a direct package file. Compare the source page and any extracted docs with my current installation, then summarize what changed and what manual follow-up I still need. Then review README.md for any prerequisites, environment setup, or post-install checks.
Trust & source
Release facts
- Source
- Tencent SkillHub
- Verification
- Indexed source record
- Version
- 2.1.0
Provenance
- Publisher
- Don-GBot
- Source page
- View original listing
- Canonical URL
- Open canonical page
Documentation
Metadata
name: cross-model-review version: 2.0.0 description: > Adversarial plan review using two different AI models. v2: Alternating mode β models swap writer/reviewer each round. Fully autonomous loop β no human input between rounds. Use when: building features touching auth/payments/data models, plans that will take >1hr to implement. NOT for: simple one-file fixes, research tasks, quick scripts. triggers: - "review this plan" - "cross review" - "challenge this" - "is this plan solid?" - "adversarial review"
When to Activate
Activate this skill when the user: Says any trigger phrase above Shares a plan and asks for adversarial/second-opinion review Asks you to "sanity check" a multi-step implementation plan Do NOT activate for: simple fixes, one-liners, pure research tasks.
Static Mode (v1 β backward compatible)
Fixed roles: planner always writes, reviewer always reviews. Requires human to trigger each round.
Alternating Mode (v2 β recommended)
Models swap roles each round. Fully autonomous β no human input between rounds. Flow: Round 1: Model A writes the plan. Model B reviews. Round 2: Model B rewrites (based on its own review). Model A reviews. Round 3: Model A rewrites (based on its own review). Model B reviews. ...continues alternating until both agree (reviewer says APPROVED) or max rounds hit. Why this works: Each model must implement its own critique β can't nitpick without owning the fix The other model catches over-engineering or proportionality issues Natural convergence: each round addresses the other's concerns
Autonomous Orchestration (Alternating Mode)
You (the main agent) run this loop. It's fully autonomous after kickoff.
Step 1 β Save the plan and init
node review.js init \ --plan /path/to/plan.md \ --mode alternating \ --model-a "anthropic/claude-opus-4-6" \ --model-b "openai-codex/gpt-5.3-codex" \ --project-context "Brief description for reviewer calibration" \ --out /home/ubuntu/clawd/tasks/reviews Captures workspace path from stdout.
Step 2 β The autonomous loop
while true: step = next-step(workspace) if step.action == "done": break # APPROVED! if step.action == "max-rounds": ask user: override or manual fix break if step.action == "review": spawn sub-agent with step.model, step.prompt save response to workspace/round-N-response.json parse-round(workspace, round, response) continue if step.action == "revise": spawn sub-agent with step.model, step.prompt save output plan to temp file save-plan(workspace, temp file, version) continue
Step 3 β Finalize
When the loop exits with APPROVED: node review.js finalize --workspace <workspace> Present: rounds taken, issues found/resolved, rubric scores, plan-final.md location.
CLI Reference
Commands: init Create a review workspace next-step Get next action for autonomous loop parse-round Parse a reviewer response, update issue tracker save-plan Save a revised plan version from writer output finalize Generate plan-final.md, changelog.md, summary.json status Print current workspace state init options: --plan <file> Path to plan file (required) --mode <m> "static" (default) or "alternating" --model-a <m> Model A β writes first (alternating mode, required) --model-b <m> Model B β reviews first (alternating mode, required) --reviewer-model <m> Reviewer model (static mode, required) --planner-model <m> Planner model (static mode, required) --project-context <s> Brief project context for reviewer calibration --out <dir> Output base dir (default: tasks/reviews) --max-rounds <n> Max rounds (default: 5 static, 8 alternating) --token-budget <n> Token budget for context (default: 8000) next-step options: --workspace <dir> Path to review workspace (required) Returns JSON: { action, model, round, prompt, planVersion, saveTo } Actions: "review", "revise", "done", "max-rounds" parse-round options: --workspace <dir> Path to review workspace (required) --round <n> Round number (required) --response <file> Path to raw reviewer response (required) save-plan options: --workspace <dir> Path to review workspace (required) --plan <file> Path to revised plan markdown (required) --version <n> Plan version number (required) finalize options: --workspace <dir> Path to review workspace (required) --override-reason <s> Reason for force-approving with open issues --ci-force Required in non-TTY mode when overriding status options: --workspace <dir> Path to review workspace (required) Exit codes: 0 Approved / OK 1 Revise / max-rounds 2 Error
Spawning reviewers
step = next-step(workspace) # action: "review" response = sessions_spawn(model=step.model, task=step.prompt, timeout=120s) # Save raw response to workspace/round-{step.round}-response.json parse-round(workspace, step.round, response_file) System instruction for reviewer: "You are a senior engineering reviewer. Output ONLY valid JSON matching the schema. No tool calls. No markdown fences. No preamble."
Spawning writers
step = next-step(workspace) # action: "revise" revised_plan = sessions_spawn(model=step.model, task=step.prompt, timeout=300s) # Save raw output as temp file save-plan(workspace, temp_file, step.planVersion) System instruction for writer: none needed β the prompt is self-contained.
Error handling
Reviewer timeout/failure: retry once, then ask user Writer timeout/failure: retry once, then ask user Parse error on review JSON: re-prompt reviewer once with "Your response was not valid JSON" Max rounds hit: present status to user, ask for override or manual fix
Convergence
The loop converges when the reviewer says APPROVED with no open CRITICAL/HIGH blockers. The script enforces this β if reviewer says APPROVED but blockers remain, it overrides to REVISE.
Static Mode (v1 β backward compatible)
For static mode, the original orchestration from v1 still works:
Step 1 β Init
node review.js init --plan <file> --reviewer-model <m> --planner-model <m>
Step 2 β Manual loop
For each round: build reviewer prompt from template, spawn reviewer, parse-round, revise plan yourself, continue.
Step 3 β Finalize
Same as alternating mode.
Integration with coding-agent
Before dispatching any plan to coding-agent that: Touches auth, payments, or data models Has 3+ implementation steps The user hasn't already reviewed adversarially Run cross-model-review first. Only proceed if exit code 0.
Notes
Workspace persists in tasks/reviews/ β referenceable later issues.json tracks full lifecycle of all issues meta.json stores mode, models, current round, verdict, needsRevision flag next-step is the state machine β always call it to determine what to do Dedup warnings help catch semantic drift across rounds Models must be from different provider families (cross-provider enforcement) --project-context is injected into reviewer prompts for calibration
Category context
Agent frameworks, memory systems, reasoning layers, and model-native orchestration.
Source: Tencent SkillHub
Largest current source with strong distribution and engagement signals.
Package contents
Included in package
5 Docs1 Config
- SKILL.md
- CHANGELOG.md
- CONTRIBUTING.md
- README.md
- SECURITY.md
- package.json