Kuro AI Agent MCP Server

Use when you need to check code for security issues before deployment or review.
Strongest for Solidity/EVM smart contracts (checks reentrancy, access control,
integer overflow, front-running, gas optimization). Also handles TypeScript,
Python, and general code review.

Returns a structured audit report with: vulnerability list (Critical/High/Medium/Low/Info),
suggested fixes, gas optimization tips (for Solidity), and overall risk assessment.
This is static analysis only — does not execute code or test against live contracts.
Limited to single-file audits; cannot resolve cross-file imports.

Args:
    code: The full source code to audit (paste complete file content)
    audit_type: Type of audit — "solidity" (EVM contracts), "typescript", "python", or "general" (any language)

Use when you need market intelligence on crypto, DeFi, or the agent economy.
Covers price context, key events, sentiment, and notable developments.
Especially strong on agent tokens, stablecoin infrastructure (x402, USDC),
and DeFi protocols.

Returns a structured briefing with: market overview, key events, price movements,
sentiment analysis, and outlook. Based on training knowledge — does not fetch
live prices. Best used for analysis and context, not real-time trading signals.

Args:
    focus: Focus area — "general" (full market overview), "btc" (Bitcoin), "eth" (Ethereum), "defi" (DeFi protocols), "agent" (agent tokens/economy), "stablecoins", or a specific token/protocol name

Use when assessing AI agent projects, protocols, or platforms for strategic value.
Scores on three dimensions (1-10 each) based on experience evaluating 50+ projects:
1. AI天网因素 (Skynet Factor) — Critical infrastructure for global AI network?
2. AI经济体落地 (Agent Economy) — Advances x402, ERC-8004, agent payments?
3. 未来发展潜力 (Future Potential) — Technical ceiling and growth trajectory?

Returns a structured evaluation with per-dimension scores, comparative analysis
against known projects, strengths/weaknesses, and overall rating. Opinionated
and honest — will give low scores when warranted. No side effects.

Args:
    project_name: Name of the project to evaluate
    description: What the project does, its technology stack, and stated goals
    additional_context: Optional — funding, team background, traction metrics, comparable projects

Use when you need to know what Kuro can do, how to reach Kuro via other
protocols (A2A, x402), or Kuro's identity details (DID, wallets, social).

Returns a JSON object with: name, description, capabilities list, protocol
endpoints (MCP/A2A/x402), identity (DID, wallets, Twitter, RNWY), and
substrate info. No inputs required. No side effects.

Use when you need in-depth analysis, literature review, competitive landscape,
or trend synthesis. Strongest in AI agents, crypto/DeFi, blockchain protocols,
and the agent economy (x402, ERC-8004, MCP/A2A). Also handles general tech,
business, and science topics.

Returns a markdown report with findings, analysis, key insights, and sources.
Does not access real-time data — analysis is based on training knowledge.
Cannot access paywalled content or private databases.

Args:
    topic: The research topic or question to investigate (be specific for better results)
    depth: Research depth — "quick" (1 paragraph summary), "standard" (detailed report with sections), or "deep" (comprehensive with sources and comparisons)

Use when you have long text (articles, papers, docs, threads) and need
a shorter version. Supports multiple summary styles for different audiences.

Returns a summary in the requested style. "concise" produces bullet points,
"executive" produces a business-ready paragraph, "technical" preserves key
details and data points, "eli5" uses simple language anyone can understand.
Input should be the full text to summarize (not a URL). No side effects.

Args:
    text: The full text to summarize (articles, papers, documents, conversation threads)
    style: Summary style — "concise" (bullet-point key points), "executive" (business brief paragraph), "technical" (preserves data and details), or "eli5" (simple explanation)

Use when you need accurate translation that preserves technical terminology,
especially for AI, blockchain, DeFi, and software engineering content.
Strongest pair: Chinese ↔ English. Also supports Japanese, Korean, and
other major languages.

Returns the translated text preserving original formatting (markdown, code blocks).
Handles domain-specific terms (e.g., 赛博永生→cyber immortality, 天网→Skynet, 
质押→staking) with proper context. Does not translate code syntax.

Args:
    text: The text to translate (supports markdown, plain text, or mixed content)
    to: Target language (e.g., "English", "Chinese", "Japanese", "Korean", "Spanish")
    source_language: Source language, or "auto" to detect automatically