Official agent skill

Commerce UI Tools

by anthropics in anthropics/commerce-agents

The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component.

OfficialApache-2.0Auto-check passedBackend & APIs

Install Commerce UI Tools

skills CLI
$ npx skills add anthropics/commerce-agents --skill commerce-ui-tools -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install anthropics/commerce-agents commerce-ui-tools --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/anthropics/commerce-agents.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/commerce-builder/skills/commerce-ui-tools .claude/skills/commerce-ui-tools && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
commerce-ui-tools
GitHub stars
3.2k
Token cost
~1.7k tokens
SKILL.md length
781 words
Files
1
Skills in repo
16
Repo updated
First seen
Licence
Apache-2.0

At a glance

The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component.

  • Tasks that involve Backend development
  • SKILL.md covers The contract, Suggestions, Streaming and Progressive rendering, plus 2 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Commerce UI Tools is an agent skill from anthropics/commerce-agents, published by the product's own GitHub organization. The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component. Load when building or reviewing how a commerce agent's output reaches a frontend.

Its SKILL.md is about 1.7k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Backend & APIs, covering Backend development. The repository describes itself as: Reference blueprint for building shopping and merchant agents with Claude. Examples in retail, commerce, telecom, and entertainment included. The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Backend development

Example prompts

  • “/commerce-ui-tools”

What it can do on your machine

Read from SKILL.md and the folder at commit fd4d592. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Commerce UI Tools loads about 1.7k tokens when it runs. Until then it costs about 75 tokens; SKILL.md has 781 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~75
When it runs · the whole SKILL.md, loaded when a task matches
~1.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from anthropics/commerce-agents at commit fd4d592, republished under its Apache-2.0 licence (© anthropics). 781 words, ~1,707 tokens.

Download SKILL.mdSave it as .claude/skills/commerce-ui-tools/SKILL.md (or your agent's skills folder).
name
commerce-ui-tools
description
The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component. Load when building or reviewing how a commerce agent's output reaches a frontend.

Presentation tools

Paths are in the reference repo: commerce_common/ is commerce-common/commerce_common/, shopping_agent/ is shopping-agent/core/shopping_agent/, merchant_agent/ is merchant-agent/core/merchant_agent/.

The contract

  • A component is a tool. The model's arguments carry its judgment: which ids, in what order, the reason or the note for each, what to compare on. The server joins every fact (title, price, image, metric value, change record) from records tools returned this session, so the frontend never renders a model-authored value.
  • One spec per component: PresentationComponent in commerce_common/presentation.py names the tool, the component string the host renders, the payload_model that validates the arguments (each role's tools/presentation.py), and the enrich hook (each role's enrichment.py, collected in PRESENTATION_COMPONENTS).
  • run_presentation in commerce_common/presentation.py is the one runner: validate, enrich, emit the ui event. The model reads the executor's displayed_text plus the hook's notes; the enriched payload goes to the host only.
  • An enrich hook drops ids without provenance and appends a note naming them; with nothing left it raises PresentationRefused, which comes back as a held call when it names a gate and as an error otherwise.
  • Shopping joins from ShoppingSessionState.seen_products, the cart, or an order read; merchant joins from MerchantSessionState.latest_snapshot, seen_series, seen_campaigns, seen_analyses, seen_listings, and seen_changes. enrich_change_preview also removes model text whose currency or weekday disagrees with the change record.
  • Composition rules (which component ends a turn, text carries the reasoning, cards carry the data) are static prompt material; each component's when-to-use line is its tool description (commerce-architecture's layer table).

Suggestions

  • present_suggestions is the one tool that carries the turn's chips, up to four; the model calls it in the same round as the turn's last component, or after its text when the turn has none, and a tapped chip is sent as the next user message. No other payload, built-in or extension, carries chips.
  • sanitize_suggestion_chips in commerce_common/fencing.py runs on PresentSuggestionsPayload: invisible and control characters out, whitespace collapsed, empties dropped, 80 characters each, four at most. The payload fails when no chip survives sanitizing.
  • A round of clean presentation calls that includes present_suggestions ends the turn (round_closes_turn in commerce_common/turn.py, close_on_presentation in the config): the Messages API loop stops there, and the Agent SDK runtimes stop through a PostToolBatch hook (close_on_presentation_hook in commerce_common/agent_sdk.py). Under host approval the merchant prompt says no chip approves or applies.

Streaming

The event list is the docstring of commerce_common/streaming.py; to_sse frames each event and a host ignores types it does not know. ui carries one enriched component, ui_partial the same while its call still streams, cart_update the whole cart after a cart write, and change_update a change record after it moved. outcome_events in commerce_common/turn.py stamps each ui event with its call id as stream_id and follows it with the tool_result, whose status is ok, error, or blocked. Every tool that is not a presentation tool may take a status argument first (with_status in commerce_common/execution.py): a few words for the person waiting, which the executor drops before the call runs and the runtime emits as the tool_call event's label; the web layer shows it as the activity line. The MCP servers build their tools without it. The Agent SDK runtimes run the same executor but return the payloads together in result.ui after the turn.

Show full SKILL.md (256 more words)Show less

Progressive rendering

A spec with an enrich_partial hook is rendered while its arguments arrive: the request marks its tool eager_input_streaming (with_eager_input in commerce_common/prompt_assembly.py), the orchestrator parses the buffer with parse_partial_json (a string still being written is left out with its key), calls enrich_partial (commerce_common/presentation.py), and yields ui_partial whenever partial_signature changes (StreamedRound.frame in commerce_common/turn.py; every visible change with eager_partial_frames on the config); the final ui event carries the same stream_id and replaces it. Input that streams as text that is not JSON comes back to the model as an error result and the round goes on (StreamedRound). Partial hooks are synchronous and join from session state only (partial_products, partial_comparison, partial_plan, partial_guide in shopping_agent/enrichment.py; partial_metrics, partial_digest, partial_change_preview in merchant_agent/enrichment.py).

Built-in components

  • Shopping (shopping_agent/enrichment.py): present_products, present_comparison, present_plan, present_guide, present_order_status, checkout (renders the cart; charges nothing), present_suggestions, and present_disclosure, registered only with enable_disclosures and filled from StorefrontBackend.get_disclosure.
  • Merchant (merchant_agent/enrichment.py): present_metrics, present_digest, present_change_preview, present_suggestions.

Adding a vertical component

  • Build a PresentationExtension (commerce_common/presentation.py): the tool description and input_schema the model sees, the payload_model that validates the same shape, the component name, and the enrich hook.
  • Pass it as extra_presentation_tools to ShoppingAgent or MerchantAgent (examples/travel/api/main.py); on the Agent SDK and MCP paths, subclass the toolset or pass executor_class= so the executor receives it as extensions. build_tools rejects a name that collides with a built-in, and the extension's bytes join the cached tool list (commerce-prompt-caching).
  • The seven in the repo: examples/travel/api/itinerary.py, examples/travel/api/occupancy.py, examples/telecom/api/plan_matrix.py, examples/telecom/api/plan_mix.py, examples/entertainment/api/venue_map.py, examples/entertainment/api/hold_view.py, examples/entertainment/api/event_pacing.py. Each keeps its types and its card in the vertical; nothing is added to a core package.

© anthropics, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in plugins/commerce-builder/skills/commerce-ui-tools of anthropics/commerce-agents.

Open the folder on GitHubat commit fd4d592

Compare with similar skills

Commerce UI Tools next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Commerce UI Tools compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Commerce UI Tools this skillanthropics/commerce-agents3.2k—~1.7kAutomated safety check: PassApache-2.0
Warpnet Debug BackendWarp-net/warpnet157—~2.9kAutomated safety check: PassCustom licence
Hunt Captcha Bypassmajiayu000/claude-skill-registry6661 repos~1.5kAutomated safety check: PassMIT
MCP Developmentcoollabsio/coolify63k1 repos~949Automated safety check: PassMIT
Turnstile Spincloudflare/skills3k4 repos~7.2kAutomated safety check: NotesApache-2.0
App Feedback AdminJamailar/Beav1.8k—~1.1kAutomated safety check: PassCustom licence

Similar skills

  • Warpnet Debug Backend

    Warp-net/warpnet

    A skill your agent uses when a Warpnet bug lives in the Go node/backend — the fat/member/remote node, its handlers, storage, or libp2p layer.

    157 GitHub stars~2.9k tokensUpdated yesterday
    Backend & APIsAuto-check passed
  • Hunt Captcha Bypass

    majiayu000/claude-skill-registry

    Hunt CAPTCHA Bypass — 6 distinct patterns: (1) CAPTCHA field simply omitted from the request (server-side validation absent), (2) CAPTCHA token replayed from a solved challenge (no single-use…

    666 GitHub starsUsed in 1 repo~1.5k tokens
    Backend & APIsAuto-check passed
  • MCP Development

    coollabsio/coolify

    A skill your agent uses for Laravel MCP development. An agent skill from coollabsio/coolify.

    63k GitHub starsUsed in 1 repo~949 tokens
    Frontend & DesignAuto-check passed
  • Turnstile Spin

    cloudflare/skills

    Official

    Set up, repair, or migrate to Cloudflare Turnstile bot verification in an existing frontend and backend, including server-side Siteverify.

    3k GitHub starsUsed in 4 repos~7.2k tokens
    Frontend & DesignAuto-check: notes
  • App Feedback Admin

    Jamailar/Beav

    Read platform-admin feedback issues through the trusted HTTP admin API, summarize user reports, and turn them into RedBox/RedConvert repair and iteration plans.

    1.8k GitHub stars~1.1k tokensUpdated today
    Frontend & DesignAuto-check passed
  • Flycms Dev

    sunkaifei/FlyCms

    FlyCms 项目(backend/ Spring Boot 4.1.1 + frontend/ vue-vben-admin v5)的架构地图与开发规范总纲。凡在本仓库做任何开发——写后端接口、新增/修改模块、管理页面、数据库变更、修 bug、重构——都要先加载本 skill 再动手,即使用户只说"改一下""加个功能";前端登录/菜单/权限专项另见…

    656 GitHub stars~827 tokensUpdated 6 days ago
    Backend & APIsAuto-check passed

More from anthropics/commerce-agents

All 16 skills in this repo
  • Commerce Evals

    anthropics/commerce-agents

    Official

    Authoring and running behavioral evals for a shopping or merchant agent, covering the case shape, authoring rules, code graders and judges, the run pattern, and poisoned fixtures.

    3.2k GitHub stars~1.8k tokensUpdated 6 days ago
    Auto-check passed
  • Catalog Listings

    anthropics/commerce-agents

    Official

    Creating and improving listing content, covering titles, descriptions, attribute completeness, categorization fixes, image callouts written as text, edits written from material the operator…

    3.2k GitHub stars~1k tokensUpdated 6 days ago
    Auto-check passed
  • Commerce Architecture

    anthropics/commerce-agents

    Official

    How the reference commerce agents of either role are structured, covering the loop, where each rule lives, skills, the backend interface, delegates, and model fields.

    3.2k GitHub stars~1.7k tokensUpdated 6 days ago
    Auto-check passed
  • Commerce Merchant Operations

    anthropics/commerce-agents

    Official

    The reference merchant agent, covering its flows, staged changes and host approval, metrics grounding, the analysis delegate, store memory, components, and marketplace rules.

    3.2k GitHub stars~2.1k tokensUpdated 6 days ago
    Auto-check passed
  • Commerce Prompt Caching

    anthropics/commerce-agents

    Official

    The reference agents' cache-stable request assembly, covering the static system and per-request context split, the fixed tool list, the rolling conversation breakpoint, which config fields are…

    3.2k GitHub stars~2k tokensUpdated 6 days ago
    Auto-check passed
  • Commerce Trust Safety

    anthropics/commerce-agents

    Official

    The rules the reference agents enforce in code for third-party content, writes, grounding, identity, and memory, each with its module, plus two adversarial-eval rules.

    3.2k GitHub stars~1.8k tokensUpdated 6 days ago
    Auto-check passed

Questions about Commerce UI Tools

What does Commerce UI Tools do?

The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component. Commerce UI Tools is an agent skill from anthropics/commerce-agents, published by the product's own GitHub organization. The reference presentation-tool contract, covering server-side enrichment, suggestion chips, the event stream, progressive rendering, both roles' built-in components, and adding a vertical component.

When should I use Commerce UI Tools?

Commerce UI Tools fits situations like: tasks that involve Backend development.

How do I install Commerce UI Tools in Claude Code?

Run `npx skills add anthropics/commerce-agents --skill commerce-ui-tools -a claude-code`. Or copy the skill folder (plugins/commerce-builder/skills/commerce-ui-tools in anthropics/commerce-agents) into .claude/skills/commerce-ui-tools in your project. Claude Code loads it when a task matches its description.

How do I install Commerce UI Tools in Codex?

Run `npx skills add anthropics/commerce-agents --skill commerce-ui-tools -a codex`. Or copy the skill folder (plugins/commerce-builder/skills/commerce-ui-tools in anthropics/commerce-agents) into .agents/skills/commerce-ui-tools in your project. Codex loads it when a task matches its description.

Can I use Commerce UI Tools in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add anthropics/commerce-agents --skill commerce-ui-tools -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/commerce-ui-tools, .gemini/skills/commerce-ui-tools, .github/skills/commerce-ui-tools and .opencode/skills/commerce-ui-tools in your project.

What does Commerce UI Tools need to run?

SKILL.md names no scripts, command-line tools or credentials: Commerce UI Tools is instructions for the agent only.

Does Commerce UI Tools access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Commerce UI Tools safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Commerce UI Tools use?

Commerce UI Tools is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Commerce UI Tools use?

About 1.7k tokens (SKILL.md is roughly 6.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Commerce UI Tools?

Skills that share tags, products or a category with Commerce UI Tools: Warpnet Debug Backend (Warp-net/warpnet, 157 stars), Hunt Captcha Bypass (majiayu000/claude-skill-registry, 666 stars), MCP Development (coollabsio/coolify, 63k stars) and Turnstile Spin (cloudflare/skills, 3k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Commerce UI Tools?

anthropics (a GitHub organization, an official publisher) maintains it in anthropics/commerce-agents, which has 3,179 GitHub stars. The repository holds 16 skills in this directory. The repository was last updated on October 2, 2026.

Source: anthropics/commerce-agents on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.