multi-llm-convergence

multi-llm-convergence is a skill for Claude Code from donnfelker/loop-skills. It costs 192 tokens per session (5,042 once invoked), scanned A, original, MIT.

A review workflow that passes any work product—such as a plan, design document, specification, or code change—between Codex and Claude reviewers until both agree it meets the chosen standard.

In plain words
What is it for?
Use it to check plans, design documents, specifications, pull requests, and implementation changes through repeated review rounds, with each round's changes recorded.
Why use it?
A single reviewer can overlook problems or repeat the same blind spots. Independent reviews help catch defects introduced or missed during earlier fixes.

Skill for Claude Code

Written for Claude Code: $CLAUDE_PLUGIN_ROOT variable. Also seen: reads .claude/ paths; mentions subagents; mentions Claude Code.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the multi-llm-convergence plugin — 1 skill shipped together

Good fit Use it to check plans, design documents, specifications, pull requests, and implementation changes through repeated review rounds, with each round's changes recorded.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add donnfelker/loop-skills
Claude Code
/plugin install multi-llm-convergence

Made for: Claude Code.

Or install multi-llm-convergence, the plugin that ships this one along with the rest of its 1 skill.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for multi-llm-convergence

README.md
[![agentmods](https://agentmods.dev/badge/skills/donnfelker/loop-skills/multi-llm-convergence/github.svg)](https://agentmods.dev/skills/donnfelker/loop-skills/multi-llm-convergence)
Your own site
<a href="https://agentmods.dev/skills/donnfelker/loop-skills/multi-llm-convergence"><img src="https://agentmods.dev/badge/skills/donnfelker/loop-skills/multi-llm-convergence/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for multi-llm-convergence

Your own site · 80×15
<a href="https://agentmods.dev/skills/donnfelker/loop-skills/multi-llm-convergence"><img src="https://agentmods.dev/badge/skills/donnfelker/loop-skills/multi-llm-convergence.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 192 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 5,042 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00192 $0.05042
Opus 5 $0.00096 $0.02521
Sonnet 5 $0.00038 $0.01008
Haiku 4.5 $0.00019 $0.00504

Measured 9d ago against content hash 4fe90a4aefab, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

multi-llm-convergence 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 9d 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.

> Why this matters: a reviewer asked to verify "does library X's function still exist" will `curl` GitHub, fail (no network → repeated `curl` exit 6), and go quiet — looking exactly like a hang. Local clones turn "fetch
plugins/multi-llm-convergence/skills/multi-llm-convergence/SKILL.md · 181 lines

How it starts

The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Multi-LLM Convergence

You are the convergence driver. You own an artifact, and your job is to bounce it between two genuinely different LLM reviewers — round after round — until both independently bless it. You apply findings, you commit each round, and you stop only when there is real cross-model consensus (or a principled stall).

Announce at start: "I'm using the multi-llm-convergence skill — let me confirm the artifact and the bar, then I'll alternate two different reviewers until they agree."

Why this skill exists

A single reviewer — even a good one — has blind spots, and a single model has correlated blind spots: ask the same model twice and it tends to miss the same things twice. Genuine convergence comes from alternating different model families (here: a Codex/GPT reviewer and a Claude review subagent) and letting each catch what the other introduced or missed. This is not theater. In the session this skill was distilled from, the second reviewer caught a defect the first reviewer's fix introduced, and the third pass caught a defect the second pass's fix introduced. Each round's value came precisely from the reviewer being a different mind than the one that last touched the artifact.

Three things make or break this loop, and all three are baked into the steps below:

  1. Ground the reviewers in local source-of-truth. Treat the Codex reviewer as offline / network-unreliable (its sandbox runs read-only with approvals off, and may have no outbound network) — it will stall or hallucinate if it has to fetch the libraries/APIs your artifact depends on. Clone them locally first.
  2. Never let a reviewer go silent. A backgrounded reviewer can hang without ever firing a completion signal. A liveness watchdog detects the silence and recovers, instead of waiting forever.
  3. Stop on real consensus, not the first "looks good." Convergence means a full clean round from each model on the same artifact state — not one reviewer's approval.

Read the full file on GitHub · 181 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 9d ago First seen · 181 lines · 192 tokens per session scan A 4fe90a4aefab

Subscribe to this mod's changes

multi-llm-convergence is a skill published in the GitHub repository donnfelker/loop-skills (18 stars, last pushed 1mo ago), licensed MIT. It adds 192 tokens to every session and 5,042 once invoked, about $0.0010 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 2026-08-30.