Adapter Author

OrganizationPopular

DefiLlama

adapter-author

Validates whether a DefiLlama request belongs in DefiLlama-Adapters, then helps coding agents create and validate TVL adapters when it does. Use when a user wants to add a new DefiLlama TVL protocol listing, check repo fit, create or fix a TVL adapter, choose adapter helpers or registries, run `node test.js`, prepare PR metadata, or decide whether they should instead use `DefiLlama/dimension-adapters` or send a mail

Overview

Publisher	DefiLlama
Repository	`DefiLlama-Adapters`
Skill name	`adapter-author`
Stars	1.3K
Forks	7.7K
Bundled files	3
Links	Skill source Repository

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.
3 bundled files
Scripts, templates, and references the model can read while it works. Files are read-only and never executed.
Open source
Published by DefiLlama on GitHub. Read the source before you install it.

Installation

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

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.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
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.

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/DefiLlama/DefiLlama-Adapters.git /tmp/DefiLlama-Adapters
mkdir -p .claude/skills
cp -r /tmp/DefiLlama-Adapters/skills/adapter-author .claude/skills/adapter-author

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Adapter Author 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 Adapter Author 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 Adapter Author 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.

DefiLlama TVL Adapter Author

Use this skill to help a protocol developer add or inspect a DefiLlama TVL adapter using repo-native patterns. The primary use case is creating new TVL adapters; narrow fixes to existing adapters are also supported when the user asks.

Step 1: Initial intake

Ask one broad question to gather context:

Tell me about the protocol, what you want added or changed on DefiLlama, why it belongs in TVL, and any chains, contracts, vaults, factories, markets, tokens, docs, oracle details, or methodology notes you already have.

After the answer, ask one unresolved question at a time. Recommend an answer when you have a defensible default. If a question can be answered by inspecting this repository or existing adapters, inspect first instead of asking. Lock each important answer before moving on to dependent decisions.

Step 2: Repository-fit gate (required before editing)

This repository (DefiLlama-Adapters) is for TVL adapters that compute TVL from on-chain data. Before editing any files, decide whether the request fits. If it does not, stop and direct the user to the correct repository:

Request type	Correct destination
Volume, fees, or revenue	`DefiLlama/dimension-adapters` (see this repo's `pull_request_template.md`)
Listing metadata only (logo, category, links)	mail with info to metadata@defillama.com
Liquidations	mail with info to metadata@defilama.com
New fetch/API-only TVL with no on-chain measurement	Stop and ask for chain-backed contracts, owners, vaults, pools, markets, or logs

API-assisted discovery is acceptable when the API only enumerates pools, vaults, markets, token lists, or config; the TVL amount itself must still come from on-chain balances, calls, or logs.

Step 3: Workflow

Read README.md, pull_request_template.md, and test.js in the repo root.
Confirm repository fit (Step 2). If the request belongs elsewhere, stop and direct the user.
If protocol facts are incomplete, follow references/protocol-intake.md and continue intake.
Classify the protocol with references/adapter-patterns.md, then open the matching example adapters and helpers before coding.
Prefer existing registry/helper patterns when the protocol cleanly matches them.
Before editing, present an understanding checkpoint to the user with:
- target path or registry entry
- protocol shape and chains
- TVL methodology and bucket classification (tvl, staking, pool2, borrowed, ownTokens)
- helper or registry choice
- unknowns and stop/ask items
- exact validation command
Edit only the adapter files needed for the chosen pattern.
Run the appropriate validation command:
- Folder adapter: node test.js projects/<protocol>/index.js
- Single-file adapter: node test.js projects/<protocol>.js
- Registry-backed adapter: node test.js <protocol-key>
Interpret the output: total TVL must exist, chain/type breakdown should be plausible, unknown-token and pricing warnings should be understood, and zero TVL must have a clear explanation.
Follow references/validation-and-pr.md to draft the PR body from pull_request_template.md. Leave unknown facts as TODO or open questions; do not guess.

Existing-adapter fixes

For changes to existing adapters, preserve the current pattern unless there is a concrete reason to change it. Inspect the current adapter, make the smallest correctness fix, run node test.js for that adapter, and revisit methodology only if the requested change affects what is counted.

Stop and ask

Stop before coding when any of the following is ambiguous: methodology, asset inclusion, double-counting, protocol-token treatment, oracle / source of truth, chain support, contract ownership, or repository fit.

Stop before declaring PR-ready when: node test.js has not passed, PR metadata is missing, unknown-token output is unexplained, package or lockfile changes are present, or the result depends on an unconfirmed assumption.

Forbidden actions

Do not add project-specific npm dependencies.
Do not edit package.json, package-lock.json, pnpm-lock.yaml, or pnpm-workspace.yaml.
Do not create fetch/API-only TVL adapters for new projects.
Do not put volume, fees, revenue, liquidations, or listing-only metadata updates in this repo.
Do not invent methodology, oracle usage, token IDs, category, audits, logo, treasury addresses, or social links.
Do not count borrowed assets as base TVL; use the lending helpers and the separate borrowed bucket when appropriate.
Do not set doublecounted, misrepresentedTokens, or permitFailure to hide uncertainty.
Do not open, push, or submit a PR unless the user explicitly asks.

Behavior examples

Good: stop a fees-adapter request and direct the user to DefiLlama/dimension-adapters before coding.

Good: use registries/erc4626.js for a plain ERC4626 vault list.

Good: use API data to discover vault addresses, then derive TVL from on-chain totalAssets.

Bad: return a third-party API's total TVL number for a new listing.

Bad: add an npm package or touch lockfiles to support one protocol.

Bad: mark something doublecounted without naming the other listed protocol whose TVL overlaps.

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 Adapter Author AI skill do?

Why use Adapter Author on TypingMind?

Because you install it once and use it with any model. Adapter Author 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 Adapter Author in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/DefiLlama/DefiLlama-Adapters/tree/main/skills/adapter-author. 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 Adapter Author?

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 Adapter Author?

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

Is the Adapter Author AI skill free?

It is published on GitHub by DefiLlama. 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.

View all

Frontend

code-yeongyu

CommunityPopular

Builds, styles, and polishes web UI and UX. Use for any frontend, page, component, styling, layout, animation, or visual-quality task, or when asked to make an interface look or feel a certain way.

Skill Catalog Builder

rohitg00

CommunityPopular

Build a bounded Agent Skill catalog across explicit discovery scopes and report collisions before loading instruction bodies.

Co Marketing

When the user wants to find co-marketing partners, plan joint campaigns, or brainstorm partnership opportunities. Use when the user says 'co-marketing,' 'partner marketing,' 'joint campaign,' 'who…

Events

When the user wants to plan, run, sponsor, speak at, or get pipeline from events — webinars, conferences, trade shows, meetups, dinners, workshops, virtual summits, or user conferences. Also use when…

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

DefiLlama TVL Adapter Author

Step 1: Initial intake

Step 2: Repository-fit gate (required before editing)

Step 3: Workflow

Existing-adapter fixes

Stop and ask

Forbidden actions

Behavior examples

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now