# Starkscan Agent Memory

- Canonical JSON memory: /.well-known/starkscan-agent.json
- Handoff JSON Schema: /agent-handoff-artifact.schema.json
- Handoff guide: /agent-workspace.md
- External API context: /api/v1/meta/agent-context
- External API capabilities: /api/v1/meta/capabilities
- External API chains: /api/v1/meta/chains
- Same-origin explorer equivalents: /v1/meta/*
- Auth header: X-Starkscan-Api-Key
- Discover the default chain from /api/v1/meta/chains

# Documentation

- [Documentation home](/docs): Starkscan explorer API, live product, and docs for Quickstart, Build, Agents, and the live API reference.
- [Quickstart](/docs/getting-started): Get connected to Starkscan quickly with a public host, a bounded API key, and no repo setup.
- [Authentication](/docs/getting-started/authentication): Send X-Starkscan-Api-Key on every request, and how to read 401, 403, and 429 responses.
- [Base URLs and chains](/docs/getting-started/base-urls-and-chains): Use /api/v1/{chain}/... for external integrations, and discover the chain from the API instead of assuming one.
- [Concepts](/docs/getting-started/concepts): Terms you will see across Starkscan docs—URLs, auth, pagination, and API tiers.
- [Get your first API key](/docs/getting-started/get-an-api-key): Create a Starkscan API key and make your first authenticated call in under five minutes.
- [Monitor 10 Wallets](/docs/getting-started/monitor-10-wallets): Canonical HTTP, SDK, and CLI starter for monitoring a fixed wallet set with recent activity, transactions, holdings, and token flows.
- [Pagination and cursors](/docs/getting-started/pagination-and-cursors): Treat cursors as opaque and follow nextCursor — never build or parse a cursor by hand.
- [Read a transaction](/docs/getting-started/read-a-transaction): Fetch a transaction's full detail and its execution trace over REST or the TypeScript SDK.
- [Your first error](/docs/getting-started/your-first-error): The handful of responses you hit on a first Starkscan call — what each means and the exact fix.
- [Build](/docs/build): Choose the right Starkscan integration surface for code and automation.
- [Launch Matrix](/docs/build/launch-matrix): Production-readiness status for Starkscan REST, RPC, SDK, CLI, and MCP surfaces.
- [Package Trust](/docs/build/package-trust): Public trust source for Starkscan npm SDK, CLI, and MCP packages.
- [API](/docs/api): How to call Starkscan REST—base URL, auth, lists, errors. Paths live in the API reference.
- [Advanced Utilities](/docs/api/advanced-utilities): External helper routes that are published, supported, and narrower than the main public lane.
- [Agent HTTP quickstart](/docs/api/agent-quickstart): Give an agent a bounded Starkscan HTTP contract without making it guess base paths, auth, or route scope.
- [Migration skills](/docs/api/migration-skills): Download source-backed SKILL.md artifacts for Voyager migration and accounting point-in-time balances.
- [Rate limits](/docs/api/rate-limits): Understand Starkscan route-class budgets, headers, and the correct retry behavior for external API keys.
- [Route Certification](/docs/api/route-certification): How Starkscan certifies REST routes, workflow evidence, and nightly correctness.
- [Route examples](/docs/api/route-examples): Exact examples for Starkscan routes that agents most often misread: block reads, search, trace, contract reads, previews, and holdings.
- [Self-serve account routes](/docs/api/self-serve): Session-authenticated routes for the signed-in user's personal Starkscan workspace and self-serve API keys.
- [SDK](/docs/sdk): First-party Starkscan SDK guidance and typed client entry points.
- [TypeScript SDK](/docs/sdk/typescript): Use the first-party Starkscan TypeScript client when you want typed explorer reads in application code.
- [Agents](/docs/ai): Use Starkscan with coding agents through the CLI and hosted MCP transport.
- [Agent CLI](/docs/ai/agent-cli): Use the Starkscan CLI for shell workflows, release validation, and local-first exports.
- [Build an agent](/docs/ai/build-an-agent): A worked example — connect the Starkscan MCP server, then explore Starknet with a safe, read-only workflow.
- [Connect your agent](/docs/ai/connect-your-agent): Add the hosted Starkscan MCP server to Claude, Cursor, Claude Desktop, Cline, VS Code, or Codex.
- [MCP Quickstart](/docs/ai/mcp-quickstart): Connect Starkscan to Codex, Claude Code, and similar MCP clients without losing the app and API context.
- [MCP tools reference](/docs/ai/mcp-tools-reference): The full Starkscan MCP tool catalog — one bootstrap tool plus 17 read-only route tools, with inputs and the REST route each maps to.
- [Reference Hub](/docs/reference): Choose the right Starkscan reference surface and use the API reference as the canonical HTTP contract.