Statewave TypeScript SDK

Official TypeScript SDK for Statewave — memory runtime for AI agents and applications.

Current release: v1.5.0 — the typed surface matches the /v1 REST contract and is semver-stable from 1.0.0 forward.

Part of Statewave — the open-source memory runtime for AI agents.

📦 Core runtime · 🐍 Python SDK · 🟦 TypeScript SDK · 🔌 Connectors · 📘 Docs · 💡 Examples · 🖥️ Admin · 🌐 statewave.ai

📋 Issues & feature requests: statewave/issues (centralized tracker)

Upgrading from a pre-0.9 SDK? v0.9.0 renamed the entire surface — request params and response fields — to idiomatic camelCase (subjectId, maxTokens, createdAt, receiptId, …). The wire protocol is unchanged; the client maps to/from the server's snake_case transparently. payload, metadata, and provenance are passed through verbatim — their inner keys are never rewritten. See CHANGELOG for the full rename table and migration steps.

New to Statewave? This SDK is a thin client for a running Statewave server. The fastest way to get one running is npx @statewavedev/statewave (or curl -fsSL https://www.statewave.ai/install | sh, or on Windows powershell -Command "irm https://www.statewave.ai/install.ps1 | iex"); see the Getting Started guide for the manual Docker Compose path. Every example below assumes a server reachable at http://localhost:8100.

Install

npm install @statewavedev/sdk

Quick start

import { StatewaveClient } from "@statewavedev/sdk";

// Basic (no auth)
const sw = new StatewaveClient("http://localhost:8100");

// With authentication and tenant
const swAuth = new StatewaveClient({
  baseUrl: "http://localhost:8100",
  apiKey: "your-key",
  tenantId: "acme",
});

// Record an episode
await sw.createEpisode({
  subjectId: "user-42",
  source: "support-chat",
  type: "conversation",
  sessionId: "ticket-8842", // optional — attribute the episode to a session (v1.0+)
  payload: {
    messages: [
      { role: "user", content: "My name is Alice and I work at Globex." },
      { role: "assistant", content: "Welcome Alice!" },
    ],
  },
});

// Compile memories (idempotent)
const result = await sw.compileMemories("user-42");
console.log(`Created ${result.memoriesCreated} memories`);

// Retrieve ranked, token-bounded context
const ctx = await sw.getContext({
  subjectId: "user-42",
  task: "Help with billing",
  maxTokens: 300,
});
console.log(ctx.assembledContext);

// Batch ingestion (up to 100)
await sw.createEpisodesBatch([
  { subjectId: "user-42", source: "crm", type: "note", payload: { text: "Prefers email" } },
  { subjectId: "user-42", source: "crm", type: "note", payload: { text: "Enterprise plan" } },
]);

// Search memories
const facts = await sw.searchMemories({
  subjectId: "user-42",
  kind: "profile_fact",
});

// Semantic search (requires embeddings)
const results = await sw.searchMemories({
  subjectId: "user-42",
  query: "billing",
  semantic: true,
});

// List all known subjects
const subjects = await sw.listSubjects();
for (const s of subjects.subjects) {
  console.log(`${s.subjectId}: ${s.episodeCount} episodes, ${s.memoryCount} memories`);
}

// Get timeline — the server's default page is the OLDEST episodes
const timeline = await sw.getTimeline("user-42");
console.log(`${timeline.episodes.length} episodes, ${timeline.memories.length} memories`);

// A bounded window of RECENT history (ascending within the page)
const recent = await sw.getTimeline("user-42", { limit: 20, newestFirst: true });
if (recent.episodesHasMore) console.log("older episodes exist beyond this page");

// Delete all subject data
await sw.deleteSubject("user-42");

Governance & audit (v0.8+)

The SDK surfaces the state-assembly receipts and sensitivity-labels / policy layer added in server v0.8, plus the v0.9 HMAC signing and as-of replay surfaces.

import { StatewaveClient } from "@statewavedev/sdk";

const sw = new StatewaveClient({
  baseUrl: "http://localhost:8100",
  apiKey: "your-key",
  tenantId: "acme",
});

// Per-request opt-in for an immutable audit receipt of the assembly.
// callerId / callerType feed the sensitivity-label policy engine —
// when the tenant config sets require_caller_identity=true, missing
// values 401.
const bundle = await sw.getContext({
  subjectId: "user-42",
  task: "What plan is this customer on?",
  emitReceipt: true,
  callerId: "agent-7",
  callerType: "support_agent",
});

if (bundle.receiptId) {
  // Receipts are ULID-addressable, tenant-scoped, append-only.
  const receipt = await sw.getReceipt(bundle.receiptId);
  // output.contextHash is a SHA-256 of the bytes delivered to the
  // agent — recompute from bundle.assembledContext to verify integrity.
  console.log(receipt.output.contextHash);
  console.log(`${receipt.selectedEntries.length} entries influenced this bundle`);
}

// List receipts for a subject, cursor-paginated, newest-first.
const { receipts, nextCursor } = await sw.listReceipts({
  subjectId: "user-42",
  limit: 10,
});
for (const r of receipts) {
  console.log(r.receiptId, r.task);
}

// Verify the HMAC signature on a stored receipt (v0.9+).
// `valid` is true | false | null — see ReceiptVerifyResult for the
// full reason vocabulary (no_signature / key_unavailable / etc.).
if (bundle.receiptId) {
  const verdict = await sw.verifyReceipt(bundle.receiptId);
  if (verdict.valid === true) {
    console.log(`signature OK — signed by ${verdict.keyId}`);
  } else if (verdict.valid === false) {
    console.log("signature mismatch — body may have been tampered with");
  } else {
    console.log(`verdict undetermined: ${verdict.reason}`);
  }
}

// Replay the receipt against current memories using the original
// policy bundle captured on the receipt (v0.9+). Returns a diff
// envelope showing what changed since emission. Pre-v0.9 receipts
// throw StatewaveUnreplayableError with reason="missing_policy_snapshot".
import { StatewaveUnreplayableError } from "@statewavedev/sdk";
try {
  const replay = await sw.replayReceipt(bundle.receiptId!);
  if (replay.diff.contextHash.changed) {
    console.log(`replay differs from original: new id ${replay.replayReceiptId}`);
  }
} catch (err) {
  if (err instanceof StatewaveUnreplayableError) {
    // err.reason ∈ {"missing_policy_snapshot", "nested_replay", "invalid_snapshot"}
    console.log(`replay refused: ${err.reason}`);
  } else {
    throw err;
  }
}

// Set per-memory sensitivity labels (server normalizes — dedup, lowercase, trim).
// Memories with labels become subject to any active policy bundle for the tenant.
const updated = await sw.setMemoryLabels({
  memoryId: "mem-uuid",
  sensitivityLabels: ["pii", "financial"],
});
console.log(updated.sensitivityLabels); // → ["financial", "pii"]

Receipts and the policy engine cooperate: every assembly call records its policy decisions into receipt.policy.filtersApplied (one entry per memory the policy fired on) and receipt.policy.filtersSkipped (per-rule summary of what didn't fire). In log_only mode (the tenant default) the receipt is the full audit trail without filtering; under enforce denied memories are dropped before they reach the assembly and the deny is still recorded. See receipts.md and sensitivity-labels.md for the full schemas and policy YAML format.

Support-agent endpoints

Statewave's support wedge — customer health scoring, SLA tracking, resolution state, and structured escalation briefs — is exposed through ergonomic SDK methods (server v0.6+).

import { StatewaveClient } from "@statewavedev/sdk";

const sw = new StatewaveClient("http://localhost:8100");

// Customer health score (0–100) with the explainable factors behind it.
const health = await sw.getHealth("customer:globex");
console.log(`${health.score}/100 — ${health.state}`);
for (const f of health.factors) {
  console.log(`  ${f.signal}: ${f.impact >= 0 ? "+" : ""}${f.impact} (${f.detail})`);
}

// SLA metrics — first-response / resolution times and breach counts.
// Thresholds are optional; they default server-side to 5 min / 24 h.
const sla = await sw.getSLA({
  subjectId: "customer:globex",
  firstResponseThresholdMinutes: 10,
  resolutionThresholdHours: 48,
});
console.log(`${sla.resolvedSessions}/${sla.totalSessions} resolved, ${sla.resolutionBreachCount} SLA breaches`);

// Track resolution state for a session (upserts by subject + session).
await sw.createResolution({
  subjectId: "customer:globex",
  sessionId: "ticket-8842",
  status: "resolved",
  resolutionSummary: "Issued refund for the duplicate charge",
});

// List resolutions, optionally filtered by status.
const openItems = await sw.listResolutions({
  subjectId: "customer:globex",
  status: "open",
});

// Generate a handoff context pack for escalation or shift change.
// `handoffNotes` is a pre-rendered markdown brief for human or LLM use.
const handoff = await sw.createHandoff({
  subjectId: "customer:globex",
  sessionId: "ticket-8842",
  reason: "escalation",
  callerId: "agent-7",
  callerType: "support_agent",
});
console.log(handoff.handoffNotes);

getHealth, getSLA, createResolution, listResolutions, and createHandoff respect the same auth, tenant-scoping, and retry behaviour as the rest of the client. createHandoff shares getContext's caller-identity gate — when the tenant config sets require_caller_identity: true, both callerId and callerType are mandatory.

Error handling

import { StatewaveClient, StatewaveAPIError, StatewaveConnectionError } from "@statewavedev/sdk";

const sw = new StatewaveClient();

try {
  await sw.compileMemories("user-42");
} catch (e) {
  if (e instanceof StatewaveAPIError) {
    console.error(`API error [${e.statusCode}]: ${e.code} — ${e.message}`);
    console.error(`Request ID: ${e.requestId}`);
  } else if (e instanceof StatewaveConnectionError) {
    console.error("Cannot connect to Statewave server");
  }
}

Where does data go?

The SDK is a thin client over the Statewave HTTP API. What leaves the network is determined by the server's compiler and embedding configuration, not by the SDK:

• Default deployment (heuristic compiler, no embeddings) — nothing leaves your infrastructure.
• LLM compiler or hosted embeddings — the server sends content to the provider you configure.

See Privacy & Data Flow for the full breakdown.

Types

All response types are fully typed:

• Episode — raw interaction record
• Memory — compiled memory with provenance + optional sensitivityLabels
• CompileResult — compilation response
• SearchResult — search response
• ContextBundle — assembled context with facts, episodes, provenance, optional receiptId / receiptEmitted
• Timeline — chronological subject history
• DeleteResult — deletion confirmation
• BatchCreateResult — batch ingestion response
• SubjectSummary — subject with episode/memory counts
• ListSubjectsResult — paginated subject listing
• Receipt + ReceiptSelectedEntry + ReceiptPolicy + ReceiptOutput — state-assembly audit artifact (v0.8+) and its nested shapes; v0.9 added HMAC signature fields (receiptSignatureKeyId, receiptSignatureAlgorithm), policySnapshot for replay, and region for residency
• ReceiptVerifyResult — valid (true | false | null) + keyId + algorithm + reason for the v0.9 HMAC verify endpoint
• ReceiptReplayResult / ReceiptReplayDiff — original + replay receipt ids plus the structural diff envelope from POST /v1/receipts/{id}/replay (v0.9)
• StatewaveUnreplayableError — thrown by replayReceipt(...) on HTTP 422; .reason is a discriminated union of "missing_policy_snapshot" | "nested_replay" | "invalid_snapshot"
• ReceiptList — cursor-paginated receipt listing
• Health + HealthFactor — customer health score and its explainable factors
• SLASummary + SessionSLA — SLA metrics, aggregate and per-session
• Handoff + ResolutionSummaryItem — handoff context pack and its prior-resolution items
• Resolution — resolution tracking record
• HealthState / ResolutionStatus — string-literal status unions

Param types: CreateEpisodeParams, SearchMemoriesParams, GetContextParams, ListReceiptsParams, SetMemoryLabelsParams, GetSLAParams, CreateHandoffParams, CreateResolutionParams, ListResolutionsParams

CreateEpisodeParams accepts an optional sessionId (v1.0+) to attribute an episode to a session; it is forwarded on both createEpisode and createEpisodesBatch.

Running tests

npm install
npm test

License

Apache-2.0