Skip to main content

November 2026

1

Nov — Watch conversations live, human handover, and file attachments

Messaging API · Management API · SDK 3.4.1–3.4.2
  • Watch a conversation live (3.4.2) — keep a WebSocket open, send a watch frame, and every message the conversation saves is pushed to you: a reply, an answer from a person on your team, the agent speaking after an outside approval. In the SDK, socket.sendWatch().
  • Human handover (3.4.2) — on a conversation the agent handed to a person on your team, every message says what happened to it in message.disposition, and a person’s message carries their name in sender.name. Your backend sends the person’s side as handover events.
  • Live cards in streams (3.4.2) — a cards event ahead of the terminal message carries the turn’s cards so far, so you can show them while the reply is still streaming.
  • File attachments (3.4.1) — create an upload, PUT the file, and attach it to a message with file_urls. In the SDK, messaging.createUpload().
  • Usage (3.4.2) — usage responses add cache_write_tokens and incomplete_usage_calls.
  • Upgrading the SDK — 3.4.2 changes nothing at runtime. In TypeScript, a socket handler that reads seq needs a 'seq' in frame guard, and an exhaustive switch needs the new watching, silent, and cards cases.
See Watch a conversation live, Human handover, and Attach files.
2

Nov — Release notes, update notices, and fault reports

Apollo CLI 0.1.109–0.1.113
  • Report a fault (0.1.112) — apollo report files an engine, CLI, platform, or docs fault with its evidence, and the CLI files its own crashes and 5xx failures in the background. APOLLO_REPORTS=0 turns both off. 0.1.113 attaches more of what happened, where, and when.
  • What changed, and what to do (0.1.111) — apollo release-notes says what changed in the CLI and in the runtime your checkout runs on, and what to do after a version change. After an upgrade, apollo upgrade names the --since command to read.
  • Told when you fall behind (0.1.110) — commands now say when a checkout’s skills, AGENTS.md, and schemas came from an older runtime than the one its version runs on, or when a newer runtime is out, with the command to run. apollo skills update refreshes them without touching the program. APOLLO_UPDATE_CHECK=0 turns the requests off.
  • Agents stay in your organization (0.1.109) — apollo agent create refuses a project outside the organization you’re signed in to, instead of creating the agent there.
See Staying up to date and apollo report.
3

Nov — The CLI docs catch up: live revisions, scripted chats, and more

Apollo CLIThe CLI pages now cover everything in 0.1.113, including features that shipped in earlier releases:
  • Live is one revision — apollo version live <tag> makes an exact revision live (an older one too — that’s the rollback), and pushing never moves it. lock, unlock, log, label, notes, and fetch round out apollo version; publish remains the one-step lock-and-ship.
  • Work across versions — apollo diff <ref> [other], apollo show, apollo restore, and apollo merge-from, plus apollo push --new-version when someone pushed to your version first.
  • Test like a caller — apollo chat --caller, --at, and --script with per-turn expectations; apollo thread list filters; and apollo open to read any of it in the Playground.
  • And more — apollo whoami, kb connections, kb resync, kb export, mockdb clone, reset, rows, and guide, corporate proxy support, and the build tree’s spec.yaml with its capability cards.
  • Removed flags — validate --fix, build --evaluate, and evaluate --world are gone, and regress --judge is now --explain: every regression is graded, and --explain adds the LLM judge’s explanations.
See the Command reference.
4

Nov — Try Beta from any version, under your own name

Apollo CLI 0.1.108
  • Any version can try Beta — apollo version try-beta no longer asks you to upgrade first: a version on an older runtime tries Beta as it is, and the rebuild runs on Beta.
  • Name it — --label <name> names the new version (default Beta).
  • Clear when Beta is off — while no Beta runtime is offered, the Beta commands say Beta isn’t available right now, and the TUI hides its Beta options.
See Beta runtime.

October 2026

1

Oct — Build and try agents on the Beta runtime

Apollo CLI 0.1.107The Apollo CLI and its TUI can now build and try agents on Beta, the newest runtime, whenever a Beta runtime is offered:
  • Build on Beta — apollo agent create "<name>" --beta creates the agent on Beta. It and every version made from it run there.
  • Try Beta — apollo version try-beta copies your version into a new version labelled Beta, runs it on Beta, and rebuilds and evaluates it there. Your own version doesn’t change, and --dry-run shows where it stands first.
  • BETA badges — apollo version list, apollo status --check, the TUI header, and the TUI’s Versions view mark every version on Beta.
  • In the TUI — Ctrl+K → try-beta, and Build on Beta among agent-create’s advanced options.
See Beta runtime.

September 2026

1

Sep — Traces for every agent, and inline with the reply

Messaging API · Management API · SDK 3.4.0The trace endpoints now answer for agents on the current runtime, not only runtime v1 agents:
  • One Trace, two shapes — every trace carries runtime ("v1" or "v2"). A runtime v2 trace is the runtime’s own turn trace, published as typed sections: what it understood and learned, the sources it read, the passages and citations the reply provably used (passages_used, citations_used), the policies it decided, the checks it passed, and a plain-language narrative.
  • Inline with the reply — set include_trace: true on a send, stream, or rerun and the reply carries message.trace_info; no second request.
  • Reading one runtime v2 interaction — pass the thread as well: GET /interactions/{interactionId}/trace?thread_id=… (interactionTrace(id, { thread_id }) in the SDK). Runtime v1 interactions are read by their id alone, as before.
  • Unchanged for runtime v1 agents — the same fields as before, plus runtime: "v1".
  • In the SDK (3.4.0) — Trace is now a discriminated union of RuntimeV1Trace / RuntimeV2Trace (narrow on runtime), with typed section types (RuntimeTraceFact, RuntimeTracePassage, …); include_trace on sendMessage / streamMessage / rerun; Message.trace_info; interactionTrace(id, { thread_id }) and threads.getInteractionTrace(id, { thread_id }). No other method changed.
  • Scoped to your agent — a thread of another agent answers 404 not_found on both trace endpoints.
See Traces — including how to show sources to end users — or try it with your own keys in the Apollo Agent Tester.
2

Sep — Apollo CLI: the official command line for Apollo agents

CLI stable · @aui.io/apolloThe Apollo CLI is the operator console for Apollo agents — a scriptable CLI (apollo <command>) and a full-screen terminal TUI (bare apollo) in one npm install:
  • Local-first authoring — agents check out as typed YAML programs (bundle/src/) with JSON Schemas, coding-agent skill packs, and the runtime’s AGENTS.md; validate, diff, pull, and push complete the loop.
  • Runtime testing — apollo chat talks to the live runtime, including unpushed local edits via --local ., with the full decision trace via --trace; apollo thread inspects conversations.
  • Server-side authoring — apollo advise answers grounded questions over your checkout, apollo build runs gated build turns, and apollo evaluate runs the live evaluation suite (simulated callers plus a judge).
  • Versions, knowledge, and secrets — full version lifecycle (version publish is also rollback), knowledge hubs (apollo kb), vault secrets (apollo vault), and a per-agent mock database (apollo mockdb).
  • Built for coding agents — every command is non-interactive with a stable --json envelope and predictable exit codes.
Get started: Installation.

August 2026

1

Aug — runtime v2 support across the API and SDK

API · SDKAgents can now run on runtime v2 — the current, streaming-first generation (its threads have UUID ids). Agents created before August 1 2026 stay on runtime v1. Sending is unchanged and routes automatically; the differences surface in a few places:
  • Thread reads (transcripts, get/rename/list) take a runtime_version selector — pass the runtime version the agent runs on (e.g. 0.8.0). See Threads and runtime versions.
  • SSE streams can end with a new suggestions event carrying ready-made follow-up prompts (followup_suggestions, with the thread_id and interaction_id they belong to).
  • Cards carry new optional metadata: title, capability (which agent capability produced the card), and instance (the record it’s about).
  • Streaming is over SSE — one HTTP call per turn, resumable with Last-Event-ID. WebSocket sessions stay with runtime v1; see Runtime v1 agents for the frame-by-frame mapping.
  • Sends accept an optional runtime_version build pin (advanced; normally omit).
2

Aug — SDK 3.3.5: list filters, fixed and flattened

SDK breakingManagement list methods now take their filter fields directly on the request — drop the filters: { … } wrapper (listThreads({ project_id }), listAgents(projectId, { name }), listVersions(agentId, { status })). The TypeScript compiler points at every call site.This also fixes a bug where list filters were silently not applied on the wire in 3.3.4 and earlier — from 3.3.5 they filter reliably. Requests from older SDK versions keep behaving exactly as they did.
3

Aug — API: choose the sender number on channel threads

API · SDKPOST /messaging/v1/channels/{channel}/threads now accepts an optional sender_id — the MongoDB ObjectId of a sender you connected in the Playground. Pass it to start the conversation from that connected number. Omit it to keep using the platform default sender.phone_number remains the recipient. When a specific sender is used, the response includes from so you can confirm which connected number sent the opener.See Channels and the SDK messaging client.

July 2026

1

Jul — API: structured card data (json_data)

API breakingAgent reply cards now carry the entity in two self-contained representations: rendered_jsx (unchanged — a ready-to-render JSX string for React UIs) and the new json_data — the same card as structured JSON, so you can build your own card UI in any framework (Vue, Angular, mobile).json_data.entity is a flat key–value map of the card’s fields, and json_data.sub_entities holds nested groups such as product variants.The card’s previous name, category, and parameters fields were removed — everything you need to render is in json_data.See Cards.
2

Jul — API: agent version on thread lists

APIGET /management/v1/threads items now include version_tag — the agent version the conversation ran on. Use it to tell which release of your agent handled each thread when filtering or auditing conversations.See Management → Threads.
3

Jul — API: new base URL

API breakingThe API base URL changed from https://api-v3.aui.io/apollo-api-v2 to https://api-v3.aui.io/apollo-api. All paths under it are unchanged (e.g. POST /messaging/v1/messages, POST /management/v1/auth/token).The old base URL keeps working for a few more days to give you time to migrate, and will then be removed — switch to the new one now.
4

Jul — API: agent_id removed from messaging request bodies

API breakingYour agent is now identified by your access token: when you exchange a publishable key at POST /management/v1/auth/token, the resulting token already carries the agent. Sending agent_id in the body of POST /messaging/v1/messages, POST /messaging/v1/messages/stream, or POST /messaging/v1/threads/{threadId}/rerun is no longer accepted and returns a validation error (422).Migration: remove agent_id from your messaging request bodies — responses keep the same shape. See Send messages.
5

Jul — API: channel thread initiation (WhatsApp / SMS)

API breakingThe POST /messaging/v1/channels/{channel}/threads request body changed:
  • agent_id was removed — the agent now comes from your access token.
  • user_ref_id was renamed to user_id.
  • template_id and content_variables (WhatsApp template overrides) were removed — the agent’s configured template is always used; agent_display_name remains available.
See Channels.
6

Jul — SDK v3: new clients for the Apollo API

SDK breaking · v3.2@aui.io/aui-client v3 is a full rewrite against the Apollo API. The single ApolloClient is replaced by two clients, one per credential:
  • ApolloMessagingClient — publishable key, browser-safe. Token exchange and refresh are handled internally; the agent comes from the key, not request bodies. Messaging, channels, and WebSocket sessions.
  • ApolloManagementClient — organization API key, server-side only. Projects, agents, versions, threads, and usage.
Tasks are now threads: createTask is gone (threads auto-create on the first sendMessage), task_id became thread_id, and the client defaults to the production API with no environment configuration. See the SDK upgrade guide.
7

Jul — API: threads, agent variables, welcome message & follow-up suggestions

API new
  • Update a thread — PATCH /management/v1/threads/{threadId} renames a thread (title); the full updated thread is returned.
  • Agent variables — POST /messaging/v1/messages and POST /messaging/v1/messages/stream accept an optional agent_variables object (static and dynamic) with per-message values for the agent’s configured context variables.
  • Welcome message — GET /messaging/v1/welcome-message returns your agent’s configured greeting, for opening a conversation UI before the first message.
  • Follow-up suggestions — POST /messaging/v1/followup-suggestions generates suggested next prompts from a context you provide.

March 2026

1

Mar 9 — Agent Builder API Alpha

Agent Builder API alpha · Closed to the public
  • Alpha release of the Agent Builder API — available to select partners only
  • Identity, organizations, accounts, and networks (agents) endpoints
  • Agent settings read/write: tools, parameters, entities, integrations, rules
  • Knowledge base management with file upload and website scraping
  • API Workflow integration generator