Skip to content
cccskills
FreePublished on GitHub

add-model

Add a new AI model to a provider (anthropic, openai, groq, cerebras). Use when a new model is announced and needs to be supported in aico.

The English summary is being prepared. Showing the original description.

Works withOpenAI
See how to install

Files included (2)

  • SKILL.md5.1 KB
  • metadata.json1.3 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Add Model Skill

Add a new AI model to the aico codebase.

Arguments

$ARGUMENTS should be in the format provider:model-id (e.g. anthropic:claude-opus-4-6). If only a model ID is given, infer the provider from the model name prefix:

  • claude-* → anthropic
  • gpt-*, o1*, o3* → openai
  • llama*, mixtral* → groq
  • Other → ask the user

Prerequisite: Gather Model Information

Before writing any code, gather the following from the user or from provided reference URLs:

  1. Model ID (e.g. claude-opus-4-6)
  2. Description - what the model is best at
  3. Pricing - input/output per MTok
  4. Context window and max output tokens
  5. Notable capabilities (e.g. extended thinking, vision, tool calling)
  6. Predecessor model to deprecate (optional)

If the user provides a reference URL, fetch it to extract this information.

Provider Reference URLs

When the user does not provide a reference URL, consult the official documentation for the relevant provider:

ProviderModels DocumentationPricing Documentation
Anthropichttps://platform.claude.com/docs/en/about-claude/models/overviewhttps://claude.com/pricing#api
OpenAIhttps://platform.openai.com/docs/modelshttps://openai.com/api/pricing/
Groqhttps://console.groq.com/docs/modelshttps://groq.com/pricing/
Cerebrashttps://inference-docs.cerebras.ai/models/overviewhttps://www.cerebras.ai/pricing

Fetch the relevant page(s) to extract:

  • The exact model ID string used in API calls
  • Context window size and max output tokens
  • Input/output pricing per million tokens
  • Release date and predecessor model (if applicable)
  • Special capabilities or limitations

Step-by-step Procedure

1. Create the model Go file

Create a new file at internal/providers/<provider>/<model_file_name>.go.

The file name convention uses underscores for separators:

  • claude-opus-4-6 → claude_opus_4_6.go
  • gpt-4o → gpt4o.go
  • llama-3.3-70b-versatile → llama3_3_70b.go

Use the existing model files in the same provider as a template. Each provider has a different pattern:

Anthropic models (internal/providers/anthropic/)

  • Uses github.com/anthropics/anthropic-sdk-go SDK
  • Struct has client *anthropic.Client and opts []anthropicopt.RequestOption
  • Constructor takes *anthropic.Client
  • Uses buildRequestBody() helper and m.client.Messages.New() / m.client.Messages.NewStreaming()
  • Must define a ModelName constant: const ModelNameXxx = "model-id"

Template reference: any existing claude_*.go file in the same directory.

OpenAI models (internal/providers/openai/)

  • Uses the in-house openai.APIClient
  • Constructor takes apiKey string
  • Uses BuildChatRequest() and m.client.DoPost() / m.client.DoStream()

Template reference: gpt4o.go

Groq / Cerebras models (OpenAI-compatible)

  • Uses openai.APIClient from the openai package
  • Constructor takes apiKey string
  • Delegates to openai.GenerateContent() / openai.GenerateContentStream() with the provider's Endpoint

Template reference: any model file in internal/providers/groq/ or internal/providers/cerebras/

2. Register the model in the provider file

Edit the provider's main file (internal/providers/<provider>/<provider>.go or anthropic.go):

  1. AvailableModels() - Add the new model struct to the returned slice
  2. selectModel() - Add a case for the model ID
  3. NewGenerativeModel() - Add a case for the model ID with the constructor

Place the new model at an appropriate position (typically at the top for the latest/best model).

3. Update the doc comment

Update the documentation comment block at the top of the provider file to include the new model's description and pricing.

4. Deprecate predecessor (if requested)

If the user requests deprecating an older model:

  1. Prefix its Description() return with [Deprecated] <Model Name> - superseded by <New Model>.
  2. Move its entry in the doc comment to the bottom with (Deprecated) suffix
  3. Keep all code functional — do NOT remove the model

5. Build and test

Run:

go build ./...
go test ./...

6. Verify with CLI

Run and show the user the output of:

go run ./cmd/aico models
go run ./cmd/aico models describe <new-model-id>

If a predecessor was deprecated, also show:

go run ./cmd/aico models describe <deprecated-model-id>

7. Commit

Follow the Conventional Commits format defined in CLAUDE.md:

feat(api): add <Model Name> support

If a predecessor was deprecated, mention it:

feat(api): add <Model Name> support and deprecate <Old Model>

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

[00] TnR ABSOLUTE RIGOR — Systemic Verification Doctrine. Forced cognitive exertion for high-stakes investigations. Mandates absolute deconstruction, mathematical proof chains, and ≥95% derived certainty. Zero-tolerance for heuristics or sentiment-based reasoning. Shortcuts are defined as intentional hallucinations.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

2agent

Free

Configures 2-Agent workflow between PM and implementation roles. Use when user mentions 2-Agent, 2エージェント, PM連携設定, Cursor設定, Cursor連携, 2-agent運用. Do NOT load for: 単独運用, ワークフロー実行, ハンドオフ処理.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Appends relevant hashtags to journal entries based on content analysis. Intended for LLM execution.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add a new AI coding agent to Agent Sandbox. Creates all required files (Dockerfile, templates, CI, docs) and wires the agent into the CLI, proxy, and build system.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add a new property to the AI agents database. Use when the user wants to add, create, or introduce a new column, property, field, or feature to track across all agents in the comparison matrix. Handles all four required steps - database updates, groups.json, table display, and GitHub issue templates.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add new Agent templates to the meta-agent-skills framework.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

See all skills by diegosouzapw

Report a problem with this skill