← Back to Skills Marketplace
harrylabsj

elm

by haidong · GitHub ↗ · v2.0.1 · MIT-0
cross-platform ✓ Security Clean
411
Downloads
0
Stars
3
Active Installs
7
Versions
Install in OpenClaw
/install elm
Description
Help users make better Eleme ordering decisions from public merchant and promotion information. Use when the user wants to compare Eleme stores, judge whethe...
README (SKILL.md)

elm

Use this skill when the user wants real Ele.me cart-building help, not just public promo analysis.

This is a consent-based assisted-ordering skill. It may access the user's Ele.me account, saved addresses, account-visible coupons or red packets, and cart state only after explicit user consent and only after the user has already authenticated in the official Ele.me surface. Never store or reuse addresses, account data, or session details outside the current session.

Required Inputs

Before execution, collect the minimum blocking details:

  • delivery address
  • target item or basket
  • specific requirements such as flavor, portion, no-onion, merchant preference, dietary limits, brand, or ETA preference
  • budget ceiling

If one of these is missing and it blocks execution, ask only for the missing item.

What This Skill Should Do

Default outcome:

  • resume from the user's existing authenticated Ele.me session or wait while the user logs in themselves
  • set or confirm the delivery address
  • search nearby merchants that satisfy the request
  • compare real payable totals after visible coupons, red packets, threshold discounts, delivery fee, and packaging fee
  • build the best basket under the user's requirements and budget
  • add the selected items to cart
  • tell the user the cart is ready and they should review and pay

Do not stop at generic recommendations if the user asked to place the basket.

Workflow

  1. Normalize the request.

    • Extract item, address, must-have constraints, avoid rules, budget, and urgency.
    • Convert vague language into actionable search terms and merchant filters.
  2. Access the authenticated session with consent.

    • If Ele.me is already signed in, continue.
    • If Ele.me is not signed in, open the official login surface and pause while the user logs in themselves.
    • Do not ask for, copy, store, or type login secrets or other authentication material on the user's behalf.
    • Never read out-of-band messages, inboxes, or notifications automatically.
  3. Set location and account context.

    • Confirm the delivery address or map pin before comparing merchants.
    • If multiple saved addresses exist, confirm which one should drive search and delivery pricing.
    • Use the address that best matches the user's stated delivery location.
  4. Search and shortlist nearby merchants.

    • Search the requested item using direct terms and useful synonyms.
    • Favor merchants that can actually deliver to the confirmed address.
    • Reject obvious poor fits on merchant quality, ETA, or impossible pricing.
  5. Build candidate carts.

    • Compare exact items, quantities, add-ons, and substitutes against the user's requirements.
    • If an exact match is unavailable, identify the closest acceptable substitute and ask before adding it.
    • For multi-item requests, optimize the whole basket instead of each item in isolation.
  6. Compare real checkout value.

    • Count food subtotal, visible coupons or red packets, store full-reduction offers, delivery fee, packaging fee, and any service fee shown in the UI.
    • Prefer discounts that apply naturally to the basket the user already wants.
    • If a small useful add-on unlocks a better threshold, it can be acceptable.
    • If the discount only works by padding the basket with low-value filler, call that out and avoid it.
  7. Choose the winning basket.

    • Respect hard constraints before optimizing savings.
    • Stay within budget when possible.
    • If nothing fits, present the closest fallback baskets such as cheapest acceptable and best-value stretch.
    • Use merchant quality and ETA as tie-breakers when price differences are small.
  8. Add to cart and hand off.

    • With explicit consent, add the chosen items to cart and select the best visible discount path.
    • Stop at cart or pre-payment review.
    • Tell the user exactly what is in the cart, the estimated payable total, the expected arrival time, and why this route won.

Decision Rules

Budget And Basket Fit

  • Hard requirements beat discounts.
  • If the budget is tight, keep the basket focused on the core item instead of chasing promo thresholds.
  • If the user specifies merchant, brand, or taste preferences, honor them unless they make the request impossible.

Real Savings

  • Count savings only after threshold, delivery fee, packaging fee, and visible service charges.
  • Prefer account-visible or automatically applicable coupons that genuinely lower the final payable amount.
  • If the final price edge is tiny, let delivery quality and merchant trust decide.

Merchant And Delivery Quality

Prefer:

  • merchants with believable recent reviews
  • stores with stable ETA and lower complaint risk
  • offers that still make sense without gaming the cart
  • baskets with lower downside if the order disappoints

Avoid:

  • suspiciously deep discounts with weak store credibility
  • high-fee baskets pretending to be low-price
  • poor review patterns around delays, wrong items, spills, or refund arguments

Search And Substitution

  • If the exact item is missing nearby, offer the nearest acceptable substitute before adding it.
  • For natural-language requests, translate soft intent into search heuristics instead of asking the user to restate everything in platform keywords.
  • If the user asks for multiple items, optimize the basket as one delivery decision.

Output

Use this structure unless the user asks for something shorter:

Cart Ready Status

Say whether the cart has been prepared and what still blocks completion.

Selected Merchant And Basket

List the chosen merchant, items, key constraints satisfied, and any approved substitutions.

Savings Summary

Summarize the coupon, red-packet, threshold, and fee logic that made this cart the best one.

Payment Handoff

Tell the user to review address, delivery notes, coupon selection, and then pay in the cart.

Safety Boundary

Allowed with explicit user consent:

  • using an already authenticated Ele.me session after the user logs in themselves
  • opening the official Ele.me login page and waiting for the user to complete authentication
  • choosing or confirming the delivery address
  • reading account-visible coupon, red-packet, and cart state
  • adding or removing cart items
  • selecting the best visible discount path
  • preparing checkout for the user's review

Not allowed:

  • asking the user to paste login secrets into chat for account access
  • typing or handling account authentication material on the user's behalf
  • storing addresses or account data outside the session
  • reading out-of-band messages, email, or notifications automatically
  • changing payment methods or account security settings
  • submitting payment
  • placing irreversible orders without the user taking the final payment step
  • inventing availability, ETA, or discount data that the UI does not show
Usage Guidance
This skill appears coherent and limited to building Ele.me carts after you authenticate via the official Ele.me interface. Before installing: (1) confirm you will not be asked to paste passwords or tokens into chat — the skill states it will not request login secrets and expects you to log in yourself; (2) be prepared to review the cart and complete payment manually; (3) remember the platform/agent will need the ability to open/resume your Ele.me browser session or UI — ensure you trust the environment that provides that capability; (4) if you have concerns, avoid enabling autonomous invocation or remove the skill after use. If the platform offers logs or permission controls, verify the skill’s interactions and revoke session access when finished.
Capability Analysis
Type: OpenClaw Skill Name: elm Version: 2.0.1 The 'elm' skill is an assisted-ordering tool for the Ele.me platform that translates natural language requests into shopping carts. The instructions in SKILL.md and README.md emphasize a consent-based model, explicitly prohibiting the agent from handling user credentials, storing sensitive data, or executing payments. It requires users to authenticate via official channels and focuses on legitimate tasks like merchant searching, price comparison, and cart preparation without any evidence of malicious intent or data exfiltration.
Capability Tags
cryptocan-make-purchases
Capability Assessment
Purpose & Capability
Name and description match the runtime instructions: the skill intends to resume an authenticated Ele.me session, read visible coupons/addresses/cart state, search merchants, build a basket, and add items to the cart. It does not request unrelated environment variables, binaries, or config paths.
Instruction Scope
SKILL.md keeps actions tightly scoped to Ele.me interactions (confirm address, read visible coupons, build cart, hand off for payment) and explicitly forbids asking for or storing login secrets. The instructions assume the platform/agent can open or resume an Ele.me browser/session and interact with UI state; this is reasonable but is an operational dependency rather than a security mismatch.
Install Mechanism
Instruction-only skill with no install spec or downloaded artifacts. Nothing is written to disk or fetched at install time, minimizing install-time risk.
Credentials
No environment variables, credentials, or config paths are required. The skill only states it will operate on the user's existing authenticated Ele.me session after the user has logged in, which is proportional to its purpose.
Persistence & Privilege
always:false (not forced into every agent run). The skill can be invoked autonomously by the agent (platform default), and its ability to act on an authenticated session increases potential impact if misused — but SKILL.md documents consent rules and no storage. Consider platform enforcement of those behavioral constraints.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install elm
  3. After installation, invoke the skill by name or use /elm
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.0.1
Tightened elm security boundaries so users authenticate themselves in the official Ele.me login flow before the skill resumes cart building.
v2.0.0
Upgraded elm into an assisted Ele.me ordering copilot with SMS login, address-based search, coupon-aware cart building, and payment handoff.
v1.3.0
Upgraded elm into an action-first Ele.me decision skill focused on whether hongbao and threshold promos create real savings.
v1.2.0
Security-aligned public release: low-sensitivity decision-support version without runtime or account-state behavior.
v1.1.1
Browser-commerce workflow upgrade: unified OpenClaw browser profile selection, snapshot/re-snapshot rules, commerce extraction order, and platform-specific live-page guidance.
v1.1.0
**Major Documentation Overhaul and Deprecation of Code** - SKILL.md is fully rewritten, providing in-depth strategies, optimization tips, and guidance for Eleme (饿了么) food delivery usage. - All code files and supporting documentation (SECURITY.md, SKILL_EN.md, cli.py, elm.py, requirements.txt) have been removed. - Skill now focuses on agent guidance and best practices rather than CLI tooling or browser automation. - Adds cross-platform usage info, comprehensive feature explanations, and Eleme/Meituan comparative insights.
v1.0.0
Initial release of the Eleme skill – a CLI tool for Eleme food delivery. - Search for food and restaurants with location and result limit options - Login via QR code authentication using browser automation - View available coupons and red packets - Displays delivery time, fee, rating, and sales info - Caches search results and securely stores session data locally - Requires the ecommerce-core framework
Metadata
Slug elm
Version 2.0.1
License MIT-0
All-time Installs 3
Active Installs 3
Total Versions 7
Frequently Asked Questions

What is elm?

Help users make better Eleme ordering decisions from public merchant and promotion information. Use when the user wants to compare Eleme stores, judge whethe... It is an AI Agent Skill for Claude Code / OpenClaw, with 411 downloads so far.

How do I install elm?

Run "/install elm" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is elm free?

Yes, elm is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does elm support?

elm is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created elm?

It is built and maintained by haidong (@harrylabsj); the current version is v2.0.1.

💬 Comments