multi-model-review

A review workflow that asks several independent AI models to examine the same coding plan or code change. A diff is the set of changes between two versions of code.

In plain words
What is it for?
Getting second or third opinions on plans and diffs, with read-only repository-aware reviews from cloud coding, cloud reasoning, and local models.
Why use it?
It exposes issues that one model may miss by comparing separate technical opinions, including problems caused by each model's command-line setup.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/dryvist/claude-code-plugins/multi-model-review
Any agent
npx skills add dryvist/claude-code-plugins --skill multi-model-review
Clone the repo
git clone --depth 1 https://github.com/dryvist/claude-code-plugins

Made for: Claude Code, Codex.

Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 985 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00091 $0.00985
Opus 5 $0.00046 $0.00492
Sonnet 5 $0.00018 $0.00197
Haiku 4.5 $0.00009 $0.00098

Measured 2d ago against content hash 57cbe0d7ed55, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

multi-model-review scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -sf http://localhost:<port>/v1/models | jq -r '.data[].id'
ai-delegation/skills/multi-model-review/SKILL.md · 81 lines

How it starts

The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Multi-model review

Independent model families tend to converge on the same top issues — that agreement is the signal worth the fan-out. Three invocation shapes cover most setups: a cloud coding-agent CLI (sandboxed, repo-aware), a cloud reasoning-agent CLI (general-purpose, also repo-aware), and a local model server (OpenAI-compatible HTTP endpoint). Each has its own gotcha; skipping past it wastes the whole review.

Cloud coding-agent CLI (e.g. an MCP-integrated coding agent)

  • Run it read-only: sandbox/approval flags set so it can inspect the repo but never edit it. Point it at the plan or diff path and let it read the repo itself rather than pasting the diff inline.
  • Account-pinned sessions reject explicit model overrides. A session authenticated against a subscription account (rather than a raw API key) will error on a hardcoded model id with something like "model X is not supported when using <tool> with a <provider> account." Omit the model parameter and let the session use its account default; only pass an explicit override on a session backed by a raw API key.

Cloud reasoning-agent CLI (e.g. a general-purpose agent CLI)

  • Use the CLI's non-interactive single-prompt mode with explicit directory grants for whatever it needs to read — don't rely on default working-dir scope.
  • It should be read-only by default (no auto-apply flag set); that's what makes it safe to fan a review out to without babysitting it.
  • Use the strongest reasoning tier, not the cheapest. A fast/economy tier will run, but on a factual review of current tooling it can confidently hallucinate — e.g. claiming a tool "doesn't exist" or "was deprecated" because its knowledge predates the tool's current state. Pick the provider's top reasoning tier for anything that requires up-to-date factual grounding, not just code-quality opinion.

Local model server (OpenAI-compatible)

  • Check the server is actually up before assuming a review failed. If nothing answers the local port, the inference service is stopped — start it via its own service manager (systemd unit, launchd job, container), then poll the models endpoint until it reports loaded:

Read the full file on GitHub · 81 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 81 lines · 91 tokens per session scan A 57cbe0d7ed55

Subscribe to this mod's changes

multi-model-review is a skill published in the GitHub repository dryvist/claude-code-plugins (3 stars, last pushed 2d ago), licensed Apache-2.0. It adds 91 tokens to every session and 985 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

rspec

This skill should be used when the user asks to "write specs", "create spec", "add RSpec tests", "fix failing spec", or mentions RSpec, describe blocks, it blocks, expect syntax, test doubles, or matchers. Should also be used when editing spec.rb files, working in spec/ directory, planning implementation phases that…

hoblin/claude-ruby-marketplace · 125 tokens

activerecord

This skill should be used when the user asks to "write a migration", "add a column", "add column to table", "create an index", "add a foreign key", "set up associations", "fix N+1 queries", "optimize queries", "add validations", "create callbacks", "use eager loading", or mentions ActiveRecord, belongsto, hasmany…

hoblin/claude-ruby-marketplace · 180 tokens

dragonruby

This skill should be used when the user asks to "create a game", "make a game", "game development", "dragonruby", "drgtk", "game loop", "tick method", "sprite rendering", "game state", or mentions args.outputs, args.state, args.inputs, coordinate system, collision detection, animation frames, or scene management.…

hoblin/claude-ruby-marketplace · 100 tokens

draper-decorators

This skill should be used when the user asks to "create a decorator", "write a decorator", "move logic into decorator", "clean logic out of the view", "isn't it decorator logic", "test a decorator", or mentions Draper, keeping views clean, or representation logic in decorators. Should also be used when editing…

hoblin/claude-ruby-marketplace · 131 tokens

mcp-server

This skill should be used when the user asks to "create an MCP server", "build MCP tools", "define MCP prompts", "register MCP resources", "implement Model Context Protocol", or mentions the mcp gem, MCP::Server, MCP::Tool, JSON-RPC transport, stdio transport, or streamable HTTP transport. Should also be used when…

hoblin/claude-ruby-marketplace · 98 tokens

ratatui-ruby

This skill should be used when the user asks to "create a TUI", "terminal interface", "terminal UI", "ratatui", "ratatui-ruby", "inline viewport", "full-screen terminal app", "terminal widgets", "tui.draw", "tui.pollevent", or mentions RatatuiRuby.run, managed loop, terminal rendering, Tea MVU, or building CLI…

hoblin/claude-ruby-marketplace · 119 tokens