Kleros Curate logo

Kleros Curate

OrganizationPopular
internet-court
kleros-curate

Interact with Kleros Curate registries across Ethereum Mainnet, Gnosis Chain, and Sepolia. Use when the user mentions Curate, Light Curate/LightGeneralizedTCR/LGTCR, Stake Curate/PermanentGTCR/PGTCR, Scout, token-curated registry/TCR, token lists, address/CDN tags, Goldsky, MetaEvidence, Verify Your List, or Curate calls such as addItem, removeItem, challengeItem, challengeRequest, and fundAppeal. Covers registry discovery and queries; policy/schema inspection; submissions, removals, challenges, evidence, appeals, rewards, execution; factory deployment; and frontend verification on all three networks. Also trigger for registry/list operations paired with Kleros, arbitrator, dispute, juror, or PNK context, and whenever the user names or tests kleros-curate. Do NOT use for non-Kleros registries or generic IPFS uploads without Curate context; route standalone uploads to kleros-ipfs-upload.

Overview

Publisherinternet-court
Repositoryinternet-court-skill
Skill namekleros-curate
Stars
5.8K
Forks
106
Bundled files
9
Links
Skill sourceRepository
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 9 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by internet-court on GitHub. Read the source before you install it.

Installation

Install the Kleros Curate AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/internet-court/internet-court-skill.git /tmp/internet-court-skill
mkdir -p .claude/skills
cp -r /tmp/internet-court-skill/vendored/kleros/kleros-curate .claude/skills/kleros-curate
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Kleros Curate in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Kleros Curate on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Kleros Curate is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Kleros Curate

Kleros Curate is a decentralized verification system for exclusive, policy-driven registries. Registries are governed by a policy document stored on IPFS — every submission is judged against that policy, and jurors consult it if a dispute arises.

The deposit/challenge/arbitration cycle:

  1. A submitter proposes an item and locks a submission deposit as a pledge that the item complies with the registry policy.
  2. The item enters a challenge window (configurable per registry, typically days to weeks). Anyone can review submissions during this window.
  3. If nobody challenges within the window, the item is accepted and the submitter's deposit is returned in full — compliant submissions carry no permanent inclusion cost.
  4. If a challenger believes the submission violates the policy, they lock a challenge deposit and open a Kleros dispute. Impartial Kleros jurors read the policy and render a verdict.
  5. The winner captures the loser's deposit as a bounty. The two-sided stake mechanism rewards accuracy and discourages frivolous challenges in both directions.

Why use Curate:

  • Earn by curating: users can profit by finding non-compliant entries in challengeable registries. In Stake Curate especially, successful challengers can recover their challenge deposit and win the item's stake.
  • Launch your own verification market: projects can create a Curate list for almost any verifiable standard, define their own policy, deposits, challenge windows, arbitrator/court, and governance. The result is a fast, low-overhead registry that can power a public Curate view or a custom frontend.

Why onchain-first matters:

Deposit amounts, arbitration costs, MetaEvidence URIs, and challenge windows are all live onchain state — they change when registry governors update parameters. Any cached or estimated value is a liability: an agent that submits the wrong deposit amount will have its transaction revert. Always read live values before acting.

Three contract flavors:

  • Light Curate (LGTCR) — LightGeneralizedTCR: optimistic challenge window, native-token deposits, the most widely deployed flavor. Used by the majority of Curate registries across Ethereum, Gnosis, and Sepolia.
  • Stake Curate (PGTCR) — PermanentGTCR: permanent ERC20 stake (not returned on item removal), Goldsky subgraph as primary data source, different status model (Submitted / Reincluded / Disputed / Absent
    • withdrawal flow). Identified by PGTCR-specific hallmark read calls (see references/stake-curate.md).
  • Scout — LGTCR contracts on Gnosis, specialized for 4 well-known registries (contract address tags, token lists, address tags, CDN mappings). Scout IS an overlay on LGTCR — it is not a separate contract type. Working with Scout always requires both references/scout-registries.md (Scout-specific context) and references/light-curate.md (LGTCR contract operations).

Non-negotiables

These rules apply across all Curate flavors. They are always in context; reference files may not repeat them.

  • Never guess / invent / approximate amounts, addresses, schemas, or parameters.
  • Onchain state + onchain logs are the source of truth for deposits, arbitration cost, challenge deposits, appeal status, and MetaEvidence URI.
  • Never assume a "standard token schema" — only the current MetaEvidence for that registry is authoritative.
  • Never rewrite the schema: item.json.columns must be copied verbatim from MetaEvidence; only values is dynamic.
  • Never upload or submit half-baked artifacts: no malformed JSON, broken MetaEvidence, placeholder values, unsupported field types, or unreachable policy files.
  • Never author unsupported MetaEvidence field types: for URL fields use type: "link", not url; validate every metadata.columns[].type before upload.
  • Production registries need a logo: do not deploy a production list with missing metadata.logoURI.
  • Strongly prefer PDF policy documents for registry policies. Use a non-PDF policy only after the user explicitly accepts the review and compatibility risk.
  • Never include "typical ranges" or estimates for deposits or fees — only report live-read values.
  • eth_getCode before declaring any address is or isn't a contract.

Which Curate flavor are you using?

Step 1 — Keyword scan (zero cost)

  • Mentions "Verify Your List", list verification, list-of-lists, frontend visibility, or making a deployed registry discoverable → Verification target is Light Curate (LGTCR) on all three supported networks → Read references/verify-your-list.md AND references/light-curate.md; do not ask the generic flavor question for the verification target. Detect the flavor of the registry being listed separately only when the live verification policy makes it relevant.

  • Mentions "Scout", "token list", "address tags", "CDN" → Scout (overlay on Light Curate — LGTCR contracts on Gnosis) → Read references/scout-registries.md AND references/light-curate.md (both required; Scout adds context on top of LGTCR operations)

  • Mentions "PGTCR", "Stake Curate", "PermanentGTCR", "Goldsky" → Stake Curate (PGTCR) → Read references/stake-curate.md

  • Mentions "Curate", "LGTCR", "LightGeneralizedTCR", "Light Curate", "addItem", "registry", or no flavor hint → Light Curate (LGTCR) (also the default) → Read references/light-curate.md

Step 2 — Ambiguous ("Curate" with no flavor hint)

  • Interactive session: ask one question — "Which Curate flavor? Light Curate (optimistic challenge window, native-token deposits), Stake Curate (permanent ERC20 stake, Goldsky subgraph), or Scout (4 Gnosis registries for contract/token tagging)?"

  • One-shot / non-interactive: default to Light Curate, then progressively correct:

    • If user provides a contract address: check if it matches one of the 4 known Scout registry addresses → load Scout overlay (read references/scout-registries.md AND references/light-curate.md)
    • If contract introspection reveals PGTCR hallmarks → pivot to Stake Curate (read references/stake-curate.md for hallmark calls)
    • Otherwise: proceed as Light Curate

Step 3 — Contract-type verification (if address provided)

See the flavor reference file for hallmark calls — SKILL.md does not embed function signatures here (contract-type detection belongs in each flavor's reference file).

Action index

Submit item to a registry → references/light-curate.md (LGTCR) or references/stake-curate.md (PGTCR)

Challenge / remove an item → flavor reference file

Submit evidence → flavor reference file

Fund an appeal → flavor reference file

Deploy a new registry (factory) → flavor reference file (factory section)

Fetch MetaEvidence (policy + schema) → references/shared-metaevidence.md

Compute deposits → references/shared-deposits.md

Build item.json → references/shared-item-json.md

Verify / make a deployed list visible on the Curate frontend -> references/verify-your-list.md

Upload to IPFS → references/shared-ipfs-upload.md

ABI / function signatures → references/shared-abi-fragments.md grep: grep -n "function\|event" references/shared-abi-fragments.md

Scout registry addresses + seed templates → references/scout-registries.md grep: grep -n "0x\|ATQ\|Address Tags\|Tokens\|CDN" references/scout-registries.md

Common workflows

Submit an item (any flavor):

  1. references/shared-metaevidence.md — fetch schema (columns) and policy URI
  2. references/shared-item-json.md — build the item.json payload
  3. references/shared-ipfs-upload.md — upload item.json to IPFS, get CID
  4. references/shared-deposits.md — compute exact msg.value
  5. Flavor reference (light-curate.md or stake-curate.md) — send the addItem transaction

Challenge or remove an item:

  1. references/shared-metaevidence.md — fetch the applicable policy (clearing policy for removal; registration policy for challenge)
  2. references/shared-ipfs-upload.md — upload evidence JSON to IPFS
  3. references/shared-deposits.md — compute the challenge deposit
  4. Flavor reference — send the challenge or removeItem transaction

Deploy a new registry:

  1. references/shared-metaevidence.md - prepare valid MetaEvidence JSON (policy URI + column schema + logoURI)
  2. references/shared-ipfs-upload.md — upload MetaEvidence JSON to IPFS
  3. references/shared-abi-fragments.md - get known factory/arbitrator addresses and deploy ABI
  4. Flavor reference — call the factory deploy function
  5. references/verify-your-list.md - submit the new registry to the network's list-of-lists if frontend visibility is required

Frontend visibility after deployment:

  • Deploying a registry does not automatically make it visible on the Curate frontend.
  • Verifying a list gives it more visibility, makes it findable in the frontend, and marks it as a listed registry for users.
  • List-of-lists submission is not mandatory, but it is highly recommended for public registries. Skip it only when the list is intentionally stealth/private.
  • On Ethereum Mainnet, Gnosis Chain, and Sepolia, the canonical verification registries are Light Curate. Use references/verify-your-list.md for the chain-specific addresses and verification safeguards, then follow the standard LGTCR submission workflow. The registry being listed may be another Curate flavor if the live verification policy permits it.

Reference files

These 9 files are loaded on demand — only when needed for the current task. The action index above points to the right file for each operation. Loading an unnecessary reference file wastes context.

references/light-curate.md Light Curate (LightGeneralizedTCR) operations end-to-end: minimum inputs to ask the user, registry discovery via eth_getCode and hallmark reads, MetaEvidence retrieval, item.json construction, submit item, challenge / remove item, submit evidence, fund an appeal, deploy a new registry via factory. Read this file for any LGTCR contract interaction — and always alongside references/scout-registries.md when working on Scout.

references/stake-curate.md Stake Curate (PermanentGTCR) operations end-to-end: PGTCR hallmark detection (distinguishes PGTCR from LGTCR), ERC20 approval + stake deposit flow, Goldsky subgraph as primary MetaEvidence source with onchain fallback, PGTCR status model (Submitted / Reincluded / Disputed / Absent), item withdrawal flow, deploy factory. Read this file for any PGTCR contract interaction.

references/scout-registries.md Scout-specific overlay for the 4 known Gnosis registries (Tokens / Address Tags / Contract Domain Names / CDN). Contains: the 4 registry contract addresses used for address-based routing, seed-first submission pattern (fill from existing seed template, then submit), item.json templates per registry, image guidance, incentives information. Always read alongside references/light-curate.md — Scout IS LGTCR at the contract layer; this file adds Scout-only context on top.

references/verify-your-list.md Three-network workflow for making a deployed registry visible and verified in the Curate frontend. Contains the canonical Ethereum Mainnet, Gnosis Chain, and Sepolia LightGeneralizedTCR verification addresses, keeps the transaction target distinct from the registry being listed, and delegates to the standard LGTCR MetaEvidence, item.json, IPFS, deposit, simulation, and request-finalization procedures.

references/shared-metaevidence.md Shared MetaEvidence retrieval applicable to all Curate flavors: eth_getLogs method with the correct topic0, latest applicable MetaEvidence selection, LGTCR registration-vs-clearing classification, Goldsky subgraph path for PGTCR, MetaEvidence JSON structure, policy URI extraction, and MetaEvidence authoring guardrails. Read this file before fetching MetaEvidence for any registry, regardless of flavor.

references/shared-deposits.md Shared deposit computation covering all Curate flavors: submission deposit formula for LGTCR, challenge deposit formula for LGTCR, PGTCR stake vs arbitration deposit distinction (ERC20 stake is separate from the native-token arbitration cost), arbitrationCost() read pattern, msg.value assembly rule. Read this file before computing any deposit or constructing any transaction value.

references/shared-item-json.md Strict item.json construction rules: the columns + values schema format, verbatim-copy rule for columns (copy from MetaEvidence without modification - values is the only dynamic part), GTCR field type allowlist, forbidden aliases, placeholder rejection, pre-upload validation, and NewItem event sampling to verify field order before submitting. Read this file before building any item payload for any Curate registry.

references/shared-abi-fragments.md Shared ABI fragments for all Curate contracts: LightGeneralizedTCR read and write function signatures, PermanentGTCR read and write function signatures, IArbitrator interface ABI, key event signatures (MetaEvidence, ItemStatusChange, RequestSubmitted, etc.), known LightGTCRFactory addresses, and known Kleros V1 arbitrator addresses. Use grep -n "function\|event\|address" to navigate this file. Read when you need function selectors, calldata encoding, or event topic hashes.

references/shared-ipfs-upload.md Shared IPFS upload guidance for Curate workflows: durability rationale (external pins can disappear after onchain anchoring), required recommended path via the kleros-ipfs-upload skill and Kleros x402 endpoint, /ipfs/<CID> format rule (avoid double-slash when building URLs), and explicit risk warning for any user-approved external pinning source. Read before any IPFS upload step inside a Curate workflow.

Feedback

Something broken or confusing in this skill? Report it: fetch feedback/SKILL.md — helps maintainers fix what agents silently trip over.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Kleros Curate AI skill do?

Interact with Kleros Curate registries across Ethereum Mainnet, Gnosis Chain, and Sepolia. Use when the user mentions Curate, Light Curate/LightGeneralizedTCR/LGTCR, Stake Curate/PermanentGTCR/PGTCR, Scout, token-curated registry/TCR, token lists, address/CDN tags, Goldsky, MetaEvidence, Verify Your List, or Curate calls such as addItem, removeItem, challengeItem, challengeRequest, and fundAppeal. Covers registry discovery and queries; policy/schema inspection; submissions, removals, challenges, evidence, appeals, rewards, execution; factory deployment; and frontend verification on all thre...

Why use Kleros Curate on TypingMind?

Because you install it once and use it with any model. Kleros Curate is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Kleros Curate in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/internet-court/internet-court-skill/tree/main/vendored/kleros/kleros-curate. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Kleros Curate?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Kleros Curate?

As many as you like. As long as a model supports skills, you can use Kleros Curate with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Kleros Curate AI skill free?

It is published on GitHub by internet-court. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

Related AI Skills

View all
nexu-io logo

Eng Runbook

nexu-io
OrganizationPopular

An engineering runbook — service overview, alerts table, dashboards links, common procedures with copy-pasteable commands, on-call rotation, and an incident-response checklist. Use when the brief…

11.2K
96.7K
1 files
View
addyosmani logo

Deprecation And Migration

addyosmani
CommunityPopular

Manages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production,…

10.1K
95.8K
Instructions
View
addyosmani logo

Observability And Instrumentation

addyosmani
CommunityPopular

Instruments code so production behavior is visible and diagnosable. Use when adding logging, metrics, tracing, or alerting. Use when shipping any feature that runs in production and you need evidence…

10.1K
95.8K
Instructions
View
addyosmani logo

Performance Optimization

addyosmani
CommunityPopular

Optimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load…

10.1K
95.8K
Instructions
View

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇