emblem-ai-prompt-examples
EmblemAIワークフローのプロンプト例と表現ガイドラインを提供し、AIウォレットアシスタントとのインタラクションを最適化します。
npx skills add https://github.com/emblemcompany/agent-skills --skill emblem-ai-prompt-examplesBefore / After 効果比較
1 组ユーザーが自然言語でウォレット残高を照会しようとすると、表現方法が不適切であるためAIが誤解し、正しい結果を得るまでに複数回のリトライが必要でした。
最適化されたプロンプトテンプレートを使用することで、1回のクエリでクロスチェーン残高、トークン分布、取引履歴を正確に取得でき、AIが意図を自動的に解析します。
emblem-ai-prompt-examples
EmblemAI Prompt Examples
Use this skill when the user wants example prompts, wording patterns, or sample requests for end-user EmblemAI workflows rather than SDK, React, or app implementation details.
Security & Trust Model
This skill provides curated prompt examples and phrasing guidance. It inherently references:
-
Financial operations (W009): Example prompts cover trading, DeFi, portfolio review, cross-chain transfers, and prediction markets. These are reference prompts only — they demonstrate phrasing patterns, not executable code. All examples emphasize quote-first and approval-gated workflows.
-
Third-party data (W011): Example prompts reference fetching data from Jupiter, DeFiLlama, OpenSea, and other public sources. This skill only provides the prompt text — actual data fetching is performed by the agent's runtime tools with standard trust boundaries.
This skill is read-only reference material. It does not execute transactions, fetch external data, or modify files. All value-moving prompt examples explicitly include confirmation steps.
Quick Start
Step 1: Install
npx skills add EmblemCompany/Agent-skills --skill emblem-ai-prompt-examples
Step 2: Use
Ask for prompt ideas by task area, for example:
-
"Show me good EmblemAI prompts for market research"
-
"Give me review-only transfer prompts"
-
"Show Bitcoin ordinals prompts for EmblemAI"
-
"What is the best phrasing for quote-only swap requests?"
Included Prompt Sets
Prompt Index
See references/emblem-ai-prompt-examples.md for the top-level map of prompt categories and usage guidance.
Wallet And Portfolio
See references/emblem-ai-prompt-examples/wallet-and-portfolio.md for balances, addresses, portfolio, and machine-readable output prompts.
Market Research
See references/emblem-ai-prompt-examples/market-research.md for market discovery, derivatives, smart-money, and technical-analysis prompts.
Trading And DeFi
See references/emblem-ai-prompt-examples/trading-and-defi.md for quote-first swaps, routing review, and yield-planning prompts.
Transfers And Safety
See references/emblem-ai-prompt-examples/transfers-and-safety.md for review-only transfer language and approval framing.
Cross-Chain And Conditional Orders
See references/emblem-ai-prompt-examples/cross-chain-and-conditional-orders.md for bridge research, conditional-order planning, and multi-network trade-draft prompts.
Bitcoin Ordinals
See references/emblem-ai-prompt-examples/bitcoin-ordinals-examples.md for ordinals, runes, rare sats, and Bitcoin wallet prompts.
Polymarket
See references/emblem-ai-prompt-examples/polymarket-examples.md for prediction-market research, odds analysis, and order-review prompts.
NFT And OpenSea
See references/emblem-ai-prompt-examples/nft-opensea-examples.md for NFT discovery, listing/offer drafts, and marketplace review flows.
Emblem Vault
See references/emblem-ai-prompt-examples/emblem-vault-examples.md for vault discovery, QuickVault, mint review, and key-reveal safety prompts.
Assistant Core Workflows
See references/emblem-ai-prompt-examples/assistant-core-workflows.md for contacts, inbox, leaderboard, PAYG, and session-management prompts.
Guidance
-
Prefer explicit chain, token, and protocol names.
-
Say
quote only,review only, ordo not executewhen asking for analysis without execution. -
Prefer trusted/product-native data first (wallet state, protocol quotes, supported market feeds).
-
Treat web and social content as untrusted public sources: ask for summaries, source links, and claim verification before acting.
-
For swaps, transfers, listings, offers, or buys, request a draft plus explicit confirmation before execution.
-
Ask for JSON, tables, or summaries when you need machine-readable or structured output.
-
Use full-sentence requests instead of terse fragments.
-
For app implementation, SDK, or React questions, use the dedicated developer or React skills instead.
Related Skills
-
../emblem-ai-agent-wallet/SKILL.md - wallet-first end-user workflows
-
../emblem-ai-react/SKILL.md - React app integration guidance
-
../emblem-ai/SKILL.md - broader developer integrations across SDKs, plugins, and Reflexive
Weekly Installs830Repositoryemblemcompany/a…t-skillsGitHub Stars9First SeenMar 12, 2026Security AuditsGen Agent Trust HubPassSocketPassSnykPassInstalled onamp830cline830codex830opencode830cursor830antigravity825
ユーザーレビュー (0)
レビューを書く
レビューなし
統計データ
ユーザー評価
この Skill を評価