Solana Agent Kit
Connect any AI agent to Solana protocols with 60+ actions.
Evidence shows: README requires users to provide private keys and API keys, but no least-privilege principle is documented; no user confirmation mechanism; data flow transparency is limited, only mentioning external API calls; sensitive data handling suggests users keep private keys secure but no encryption at rest; dependency security has SECURITY.md and dependency locking, but no vulnerability scanning evidence; external effects include transactions and cross-chain bridges, but no risk warnings; rollback mechanism not mentioned; source attribution: author is sendaifun, but not verified. Deductions: lack of user confirmation, rollback, and least-privilege practices.
Evidence shows: README and code examples are consistent, but there are inconsistencies, e.g., missing comma in create3LandSingle example; dependency availability: many dependencies, no lockfile provided; failure messages: code may have error handling, but no specific examples. Deductions: missing lockfile, insufficient failure message examples.
Evidence shows: target audience clearly AI researchers and crypto builders, scenarios broad; capability boundaries via plugins, but specific limitations not documented; trigger precision: tool calls require precise parameters, but no parameter validation documentation; environment fit: supports multiple frameworks and Node versions, but no complete environment variable configuration. Deductions: insufficient trigger precision and boundary documentation.
Evidence shows: information architecture clear, with README, docs link, and examples; install notes detailed, including npm and plugin installation; naming stability: version 2.0.7, but no naming conventions; examples abundant, but no FAQ; known limitations not explicitly listed; license Apache-2.0, but copyright year 2024; versioning changelog not provided; maintenance responsibility: SECURITY.md and CI, but maintainers not explicit. Deductions: missing FAQ, known limitations, and changelog.
Evidence shows: output usability: provides many tools and examples, but no output format documentation; marginal value: provides 60+ actions, but no comparison with alternatives; cost-benefit: no performance or cost analysis. Deductions: lack of output format and cost-benefit analysis.
Evidence shows: README claims not backed by specific evidence, e.g., test results; cross-source corroboration: no third-party verification; fact-inference separation: not clearly distinguished. Deductions: lack of verifiable evidence.
- Private keys and API keys are exposed in code examples; users must handle with care.
- No user confirmation mechanism; automated transactions may pose risks.
- Many dependencies without lockfile; supply chain security needs attention.
What does this agent do, and when should you use it?
Solana Agent Kit is an open-source toolkit that enables AI agents to interact with Solana blockchain protocols. It provides over 60 actions including token trading, NFT creation, lending, staking, cross-chain bridging, and more. The kit integrates with LangChain and Vercel AI SDK, allowing agents built with any model to use blockchain tools. It features a plugin architecture with Token, NFT, DeFi, Misc, and Blinks plugins. Installation is straightforward via npm, and it includes TypeScript examples. The documentation is available at docs.sendai.fun.
The toolkit provides a SolanaAgentKit class that wraps wallet and RPC configuration. It performs token operations (deploy SPL tokens, transfer assets, check balances), NFT operations (Metaplex collection creation, 3.land minting and listing), DeFi operations (Jupiter swaps, Raydium pool creation, Drift perpetual trading, etc.), cross-chain bridging (Wormhole, deBridge), airdrops (via Light Protocol compression), and market data (CoinGecko, Pyth). It offers tool integrations via createLangchainTools, createVercelAITools, and createOpenAITools. It supports autonomous chat mode and interactive mode.
- Developers building AI trading agents that automatically swap tokens via Jupiter.
- NFT creators using AI to generate artwork and list on 3.land automatically.
- DeFi users who want to lend, borrow, and trade via natural language commands.
- Cross-chain users who bridge assets via Wormhole or deBridge without manual steps.
- Analysts fetching live token prices and trends with an AI agent.
What are this agent's strengths and limitations?
- Supports 60+ Solana actions covering multiple domains.
- Integrations with LangChain and Vercel AI SDK make it easy to wire up AI agents.
- Plugin architecture lets you load only needed features.
- Supports multiple wallet types, including embedded wallets.
- Depends on specific RPC providers and API keys (e.g., OpenAI, Helius).
- Requires Node.js environment and some blockchain knowledge.
- Complex DeFi and cross-chain actions may involve additional configuration and risk.
- Documentation is online and may lag behind the code.
How do you install or deploy this agent?
Install Node.js and npm. Run npm install solana-agent-kit, and optionally install plugins: npm install @solana-agent-kit/plugin-token @solana-agent-kit/plugin-nft @solana-agent-kit/plugin-defi @solana-agent-kit/plugin-misc @solana-agent-kit/plugin-blinks. Requires a Solana wallet private key and an RPC URL.
How do you use this agent?
Initialize SolanaAgentKit with a wallet and RPC URL. Example: const agent = new SolanaAgentKit(wallet, 'YOUR_RPC_URL', { OPENAI_API_KEY: 'YOUR_OPENAI_API_KEY' }); Then add plugins with .use(TokenPlugin), etc. Call methods like agent.methods.trade(), agent.methods.deployToken(). Create AI tools with createVercelAITools(agent, agent.actions).