claude-agent-sdk

name	claude-agent-sdk
description	Authoritative reference for working on MultiTable's Claude Agent SDK integration (`@anthropic-ai/claude-agent-sdk`). Trigger when the user mentions the agent SDK, `query()`, streaming, permission modes, `canUseTool`, `AskUserQuestion`, elicitation, hooks, abort/stop, plan mode, sessions, subagents, MCP, usage limits / rate limits (`rate_limit_event`, `rate_limit_info`), or modifying anything under `packages/daemon/src/agent/` or `packages/daemon/src/hooks/`.
allowed-tools	Read, Grep, Glob, Bash, Edit, Write

Claude Agent SDK reference for MultiTable

Pinned SDK version: @anthropic-ai/claude-agent-sdk@0.2.119 (see package.json). Type signatures in this skill are quoted from the local sdk.d.ts so they always match what we have installed — re-verify after any SDK upgrade.

This skill exists because MultiTable's daemon-side SDK integration has bitten us repeatedly on the same four things: streaming preview lifecycle, intercepting AskUserQuestion, knowing when a stream is active vs. completed, and stopping a turn mid-stream. Don't re-derive — read the relevant chapter below before changing agent/manager.ts, hooks/permissionManager.ts, or hooks/elicitationManager.ts.

Quick task → file map

What you want to do	Read
Understand the query() call we make	reference/query-and-options.md
Add or fix streaming preview behavior	reference/streaming-and-lifecycle.md + multitable/streaming-state-machine.md
Add a permission mode toggle (plan / acceptEdits / etc.)	reference/permissions-modes.md
Intercept AskUserQuestion or any other interactive prompt	reference/canusetool-and-elicitation.md + multitable/permission-and-elicitation-wiring.md
Add an SDK lifecycle hook (e.g. Notification, PreCompact)	reference/hooks.md
Resume / fork / list sessions	reference/sessions.md
Diagnose a previous_message_id 400 on resume	pitfalls.md §10 (env-var fix at top of agent/providers/claude.ts)
Stop a turn mid-stream, or fix abort regressions	reference/abort-and-stop.md
Add a custom MCP tool	reference/custom-tools-mcp.md
Use subagents	reference/subagents.md
Display or accumulate cost	reference/cost-tracking.md
Surface usage / rate limits in the per-session indicator	reference/usage-limits.md
Customize the system prompt or list slash commands	reference/system-prompts-slash-commands.md
Find where event X is emitted / forwarded	multitable/event-map.md
Get oriented in the daemon's agent code	multitable/architecture.md
Touch the Codex provider	multitable/codex-provider-notes.md
Diagnose a bug we've seen before	pitfalls.md

Decision tree: which lever to pull?

Need to BLOCK a tool call?
├── Always block, no UI? ─── Options.disallowedTools (static rule)
├── Block based on input shape (regex, file path)? ─── hooks.PreToolUse with permissionDecision: 'deny'
└── Ask the user every time? ─── canUseTool callback

Need to RUN code on a lifecycle event (turn start, tool result, compact, notification)?
└── Options.hooks  (NEVER for interactive UI gating; hooks should not block on user input)

Need the agent to ASK the user a structured question?
├── Built-in AskUserQuestion tool? ─── Intercept in canUseTool (toolName === 'AskUserQuestion')
└── MCP server elicitation/create? ─── onElicitation callback

Need to STOP the turn mid-stream?
├── From the daemon? ─── currentTurn.abortController.abort()
└── From the SDK directly? ─── Query.interrupt()  (we don't use this; we own the controller)

Need to RESUME a prior conversation?
├── Most recent? ─── Options.continue: true
├── Specific id? ─── Options.resume: claudeSessionId
└── Branch from a prior point? ─── Options.resume + Options.forkSession: true

Need PLAN mode (read-only thinking)?
└── Options.permissionMode: 'plan'  (reset per-turn, or call Query.setPermissionMode mid-stream)

Three rules that get violated most

1. The canonical assistant message is the source of truth, not the stream. Streaming assistant-delta is a preview. Always clear streaming state in the finally block of the turn — never rely on message_stop alone, because the SDK may end the iterator before sending one (abort, network drop, error).

2. canUseTool is the universal prompt-interception point for tools the agent decides to call. AskUserQuestion is just a tool name routed through canUseTool — branch on toolName === 'AskUserQuestion' and translate the structured questions into UI prompts. Never auto-defer or auto-allow it; the prompt is the whole point.

3. onElicitation is for MCP elicitation requests, not the same channel as canUseTool. It's a separate Promise-returning callback for form/url-shaped requests from MCP servers. Wire it through ElicitationManager, not PermissionManager.

Ground truth files

When in doubt about behavior, these are authoritative (in this order):

1. node_modules/@anthropic-ai/claude-agent-sdk/sdk.d.ts — the actual installed types.
2. packages/daemon/src/agent/manager.ts — how MultiTable wires the SDK.
3. docs/reference/archive/SDK_MIGRATION_PLAN.md — historical migration notes; explains why certain choices exist.
4. The Anthropic public docs (URLs cited inside each reference file). Slower to update than the type defs; cross-check before trusting.