silk
Agent banking and payments on Solana. Send and receive stablecoins with cancellable escrow transfers. Optional on-chain accounts with policy-enforced spending limits for human-delegated automation.
What this skill does
# SilkyWay Agent banking and payments on Solana. Send and receive stablecoins — non-custodial, on-chain. ## Install ```bash npm install -g @silkysquad/silk ``` Requires Node.js 18+. ## Setup ```bash # 1. Initialize (creates wallet and agent ID) silk init # 2. Check your wallet address silk wallet list ``` Your wallet and agent ID are saved at `~/.config/silk/config.json`. Your private key never leaves your machine. `silk init` is idempotent — safe to run multiple times. ### Cluster configuration Default cluster is `mainnet-beta` (real USDC). Switch to `devnet` for testing with free tokens. ```bash silk config set-cluster devnet # test tokens silk config set-cluster mainnet-beta # real USDC silk config get-cluster # show current ``` | Cluster | API Base URL | Network | |---------|-------------|---------| | `mainnet-beta` | `https://api.silkyway.ai` | Mainnet (real USDC) | | `devnet` | `https://devnet-api.silkyway.ai` | Devnet (test USDC) | ### Fund your wallet (devnet) On devnet, use the faucet — it gives you 0.1 SOL (for transaction fees) + 100 USDC: ```bash silk config set-cluster devnet silk wallet fund silk balance ``` On mainnet, send SOL and USDC to your wallet address manually. SOL is required for Solana transaction fees. ## Sending Payments ```bash silk pay <recipient> <amount> [--memo <text>] ``` This locks USDC into on-chain escrow. The recipient claims it with `silk claim`, or you cancel for a full refund with `silk cancel`. The output includes a **claim link** (`claimUrl`) — a URL you can share with the recipient's human. They open it in a browser, connect their wallet, and claim the payment. This is the easiest way for a non-technical recipient to claim. ```bash # Send 10 USDC silk pay 7xKXz9BpR3mFVDg2Thh3AG6sFRPqNrDJ4bHUkR8Y7vNx 10 --memo "Payment for code review" # Output includes claimUrl — share it with the recipient # Example: https://app.silkyway.ai/transfers/9aE5kBqRvF3...?cluster=devnet # Check your balance silk balance # View your transfers silk payments list silk payments get <transfer-pda> ``` ### Claiming a payment If someone sent you a payment: ```bash silk payments list silk claim <transfer-pda> ``` ### Cancelling a payment Cancel a payment you sent (before the recipient claims it): ```bash silk cancel <transfer-pda> ``` ## Address Book Save contacts so you can send payments by name instead of address. ```bash silk contacts add alice 7xKXz9BpR3mFVDg2Thh3AG6sFRPqNrDJ4bHUkR8Y7vNx silk contacts list silk contacts get alice silk contacts remove alice ``` Once saved, use names anywhere you'd use an address: ```bash silk pay alice 10 --memo "Thanks for the review" silk account send alice 5 ``` Contact names are case-insensitive and stored lowercase. Saved at `~/.config/silk/contacts.json`. ## Multi-Wallet Support ```bash silk wallet create second-wallet silk wallet fund --wallet second-wallet silk wallet list ``` Use `--wallet <label>` on any command to select a non-default wallet: ```bash silk pay <address> 10 --wallet second-wallet silk balance --wallet second-wallet ``` ## Support Chat ```bash silk chat "How do I send a payment?" ``` A persistent `agentId` (UUID) is auto-generated on first use for session continuity. ## On-Chain Accounts (Optional) A SilkyWay account is an on-chain wallet — like a bank account — that a human owner creates and funds with USDC. The owner can add agents as **operators** with per-transaction spending limits enforced on-chain. This is useful for automations, recurring payments, or any scenario where a human wants to delegate spending authority to an agent with guardrails. You don't need an account to use SilkyWay. Escrow payments (`silk pay`) work with just a wallet. Accounts are an optional upgrade when your human wants to give you direct spending access with on-chain controls. **Key concepts:** - **Owner** — The human who creates and funds the account. Full control: can transfer any amount, pause the account, add/remove operators. - **Operator** — You (the agent). Authorized to send tokens from the account, subject to a per-transaction limit set by the owner. - **Per-transaction limit** — Maximum USDC you can send in one transaction. Enforced on-chain — the Solana program rejects transactions that exceed it. A limit of 0 means unlimited. - **Pause** — The owner can pause the account, blocking all operator transfers until unpaused. You cannot unpause it. ### Setting up an account Your human creates the account — you cannot create it yourself. 1. Share the setup URL with your human (replace with your address from `silk wallet list`): ``` https://app.silkyway.ai/account/setup?agent=YOUR_WALLET_ADDRESS ``` They'll connect their wallet, set your spending limit, and fund the account. **Important:** Your human must select the same network (mainnet/devnet) on the setup page as your CLI cluster. If you're on devnet, tell them to switch to devnet before creating the account. 2. After your human creates the account, sync it: ```bash silk account sync ``` 3. Check your status and send payments: ```bash silk account status silk account send <recipient> <amount> ``` If the amount exceeds your per-transaction limit, the transaction is **rejected on-chain** with `ExceedsPerTxLimit`. If the account is paused, you get `AccountPaused`. If `silk account sync` returns "No account found", your human hasn't created the account yet — share the setup URL with them. ### Accounts vs escrow payments | | Accounts | Escrow (`silk pay`) | |---|---|---| | **Setup required** | Human creates account + adds you as operator | None — just a funded wallet | | **Spending limits** | Per-transaction limit enforced on-chain | No limits | | **Recipient claims?** | No — direct transfer, tokens arrive immediately | Yes — recipient must `silk claim` | | **Cancellable?** | No — transfer is instant | Yes — sender can cancel before claim | | **Best for** | Ongoing payments with human oversight | One-off payments between parties | If your human has set up an account for you, prefer `silk account send` — it's simpler (no claim step) and your human controls the spending limits. ## CLI Reference | Command | Description | |---------|-------------| | `silk init` | Initialize CLI (create wallet, agent ID, and contacts file) | | `silk wallet create [label]` | Create a new wallet | | `silk wallet list` | List all wallets with addresses | | `silk wallet fund [--sol] [--usdc] [--wallet <label>]` | Fund wallet from devnet faucet | | `silk balance [--wallet <label>]` | Show SOL and USDC balances | | `silk pay <recipient> <amount> [--memo <text>] [--wallet <label>]` | Send USDC payment into escrow | | `silk claim <transfer-pda> [--wallet <label>]` | Claim a received payment | | `silk cancel <transfer-pda> [--wallet <label>]` | Cancel a sent payment | | `silk payments list [--wallet <label>]` | List transfers | | `silk payments get <transfer-pda>` | Get transfer details | | `silk contacts add <name> <address>` | Save a contact to the address book | | `silk contacts remove <name>` | Remove a contact | | `silk contacts list` | List all saved contacts | | `silk contacts get <name>` | Look up a contact's address | | `silk account sync [--wallet <label>] [--account <pda>]` | Discover and sync your on-chain account | | `silk account status [--wallet <label>]` | Show account balance, spending limit, and pause state | | `silk account send <recipient> <amount> [--memo <text>] [--wallet <label>]` | Send from account (policy-enforced on-chain) | | `silk chat <message>` | Ask SilkyWay support agent a question | | `silk config set-cluster <cluster>` | Set cluster (`mainnet-beta` or `devnet`) | | `silk config get-cluster` | Show current cluster and API URL | | `silk config reset-cluster` | Reset cluster to default (`mainnet-beta`) | Use `--wallet <label>` on any command to select a non-default wallet. Recipients accept contact names or Solana addresses. ## How
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.