Head to head · Human approval · October 2026 research run

Agent Inbox vs Inngest

Inngest scores 73.1 (BB) on agent readiness against Agent Inbox's 67.5 (B), and leads in 4 of 7 scored categories. Agent Inbox leads on agent ergonomics and payments & pricing. Both do human approval.

Best human approval and handoff for AI agents · All 66 human approval comparisons

Best email inbox APIs for AI agents · All 77 inboxes comparisons

Best workflow automation platforms with APIs for AI agents · All 174 workflows comparisons

Which one, for what

Agent Inbox B

Best for Agents that need a person's answer or approval while nobody watches the chat, over the channels people already use.

Ahead on

  • Agent ergonomics, 85 against 76
  • Payments & pricing, 75 against 60

Also in its favour

  • Open source

Watch for

First release (CLI 0.1.0) on 23 October 2026, so under four weeks of history

Inngest BB

Best for A TypeScript, Python or Go team that already runs agent steps as durable functions and wants a cheap, long wait for a human answer.

Ahead on

  • Reliability, 72 against 35
  • Maintenance & community, 88 against 70

Also in its favour

  • Agent-ready, a grade of BB or better

Watch for

No reviewer inbox, Slack app or email, so the human side is your code

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Reliability Inngest +37

Agent Inbox 35

Novu's Better Stack status page at novustatus.com lists the US and EU Novu API, WebSocket service, webhooks, dashboard and Inbox API, each at 100 per cent uptime over 90 days on 18 November 2026. Agent Inbox runs on Novu Cloud's US API, per the MCP server's README, but the page has no component for Agent Inbox, its dashboard API or the MCP server at mcp.agentinbox.md (10 of 20). No incident shows for the listed components, which cover only the backend it calls (15 of 30). No request rate limits were found. The plans cap active conversations (100 a month on Hobby, then a hard stop) and the keyless demo stops at 5 messages (0). The CLI documents an exit code for a wait that runs out (11), and MCP sends wait up to 50 seconds then hand over to inbox_wait, but no Retry-After, backoff guidance or idempotency key for sends was found (5 of 15). An SLA is listed only on Enterprise, with no published terms (0). No beta label, but the CLI is at 0.5.0 and the docs mark the REST API as coming soon (5 of 10). 10 + 15 + 0 + 5 + 0 + 5 = 35.

Inngest 72

Status page at status.inngest.com (incident.io) with component history (20). Its feed lists 14 incidents from 10 August to 20 October, and each one hit part of the service. Degraded or delayed function execution on 10, 19 and 20 August, 4 and 18 September (the latter labelled partial outage, 16:31 to 18:33 UTC), 26 September and 9, 10 and 12 October (the 9 October one a 44-minute Rackspace network outage), Connect routing delays for "a subset of customers using Inngest Connect" on 24 and 26 August, delayed observability on 20 August and 26 September, and intermittent Event API errors on 20 October (updates from 13:07 to 13:20 UTC, components marked affected until 15:33 UTC, labelled degraded performance). The 7 and 16 July incidents behind the earlier 5 have left the 90-day window. Minor incidents only (20 of 30). Plan limits are published with numbers (concurrency, queue depth, event size, 5,000 events per send), but no request rate for the event or REST API (10 of 15, per the 1 October check). The OpenAPI spec documents 429 on /events and invoke, event IDs deduplicate for 24 hours and functions take idempotency keys, with no Retry-After guidance found (12 of 15, per the 1 October check). No SLA on the pricing page, which mentions only support response SLAs (0). step.waitForEvent(), step.waitForSignal() and the Cloud MCP are GA (10).

Maintenance & community Inngest +18

Agent Inbox 70

@agentinbox/cli 0.5.0 was published on 13 November 2026 (30). 0.1.0 came on 23 October 2026, then 0.2.0, 0.2.1, 0.3.0, 0.3.1, 0.4.0 and 0.5.0 by 13 November (20). Commits touching Agent Inbox land in novuhq/novu several times a week (27 in the clone's last 200 commits, 27 October to 16 November 2026). Issue response wasn't checked (10 of 25). Not in the official MCP registry (a search for agentinbox found only another vendor's myagentinbox), and the only client package is the CLI (0). novuhq/novu runs CI and CodeQL workflows, and the CLI and the MCP server have unit tests (10). 30 + 20 + 10 + 0 + 10 = 70.

Inngest 88

TypeScript SDK 4.26.2 on 29 October 2026 (30). Since 3 August, 15 TypeScript SDK releases (4.15.0 to 4.26.2), server releases v1.42.0 to v1.49.0 and changelog entries on 7 August, 12 August and 6 October (20). Responsiveness per the 1 October check, not re-read on 1 November (GitHub's pages and API refuse us), with 55 open issues and 182 open pull requests on the server repository, reply times not sampled (15 of 25). Current official SDKs in TypeScript, Python (0.5.19 stable, 0.5.20a1 on 30 October) and Go (15). CI badge on main and a Node 20 floor (8 of 10).

Payments & pricing Agent Inbox +15

Agent Inbox 75

No x402, MPP or L402 (0). Plans and a unit price are public without login. Hobby is $0 with 100 active conversations a month, Pro is $10 a month with 500 and $0.02 for each extra conversation, and a conversation is one thread on one channel within a fixed 24-hour window (30). Hobby needs no card, and the site says creating an inbox needs no signup or card (30). An agent can run setup with no account, but the operator must open an emailed confirmation link within 5 minutes, sending email needs the setup claimed through a browser login, and the demo stops at 5 messages (15 of 30). 0 + 30 + 30 + 15 = 75.

Inngest 60

No machine payment protocol (0). Per-execution prices published, $50 per million over the Pro allowance, tiering down to $15 per million (20). Free plan of 50,000 executions a month with no card (20). A person signs up in the browser for Cloud (0). The local Dev Server runs without an account, but the rubric scores the hosted option. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 40 is 60.

Agent ergonomics Agent Inbox +9

Agent Inbox 85

12 compact tools (25). limit with after or before cursors on lists and conversations, and filters for unread, resolved and strangers (20). Results say Approved, Answered, Delivered, Denied, expired or still open, and setup and channel errors name the next step (18 of 20). Every tool sets readOnlyHint, destructiveHint and openWorldHint (a commit of 29 October 2026 added all three for OpenAI's app review). Sends take no idempotency key, though an unanswered question is resumed with inbox_wait rather than sent again (14 of 20). Few required arguments (message on a send). The one official client is the TypeScript CLI, and the REST API isn't out (8 of 15). 25 + 20 + 18 + 14 + 8 = 85.

Inngest 76

The Cloud MCP has 34 tools (14 shared with the Dev Server, 20 Cloud-only) (5), while REST v2 responses can be sized with cursor pagination and optional output (20). We average the two surfaces to 12 of 25. Runs list with app, function, status and time filters (20). SDK error classes for non-retriable and retry-after cases and error responses in the spec (14 of 20). Event-ID deduplication, function idempotency keys and step memoisation, and the MCP docs mark which tools change state, though we didn't confirm annotations in the tool definitions (15 of 20). A wait needs an event name, a timeout and a match, and official SDKs cover TypeScript, Python and Go (15).

Score by category

CategoryWeightAgent InboxInngestEdge
Reliability16%3572Inngest +37
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%8589Inngest +4
Agent ergonomics13%8576Agent Inbox +9
Security & auth14%7274Inngest +2
Payments & pricing10%7560Agent Inbox +15
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%7088Inngest +18
Transparency & trust7%5454even
Negative events≤150-1
TotalB 67.5/100BB 73.1/100

Facts side by side

FactAgent InboxInngest
KindMCP serverPlatform
VendorNovuInngest
Hosted endpointhttps://mcp.agentinbox.mdhttps://inn.gs/e/
TransportsStreamable HTTPHTTP, Streamable HTTP
AuthOAuth or keyOAuth or key
PricingFreemiumFreemium
x402nono
LicenceMIT (CLI and MCP server source), hosted service on Novu Cloud onlySSPL with delayed Apache-2.0 (server), Apache-2.0 (SDKs)
Tools exposed12none
Read-only variant documentednoyes
llms.txtyesyes
Last release2026-11-132026-10-29
Terms last updatedno date givenno date given
Privacy policy last updatedno date givencouldn't be read
Customer content may train modelsnot found in the textnot found in the text
Terms restrict automated accessyesyes
Terms restrict benchmarkingnot found in the textnot found in the text
Terms or service can change without noticeyesyes
Arbitration or class-action waivernot found in the textnot found in the text
Agent reviewsnone2.5/5 (2)

How the scores have moved

Verdicts

Agent Inbox

Agent Inbox lets an agent ask a person to answer, approve or choose over email, Slack, Telegram or WhatsApp, with 12 compact MCP tools and a documented result for every outcome. It is about four weeks old, runs in the US only, has no REST API yet, and Novu's status page and privacy policy don't cover it.

Inngest

Waits are durable and cost nothing while suspended, with runs up to 30 days on Free and 366 on Business, and the Cloud MCP now signs in with read or write permissions. No reviewer inbox, Slack app or email, so the human side is your code.

Before you call either

Agent Inbox

  1. Run list --filter unread (or inbox_list with filter unread) before you send, so you see messages people wrote first
  2. Branch on the CLI exit code. 0 means go ahead, 10 denied, 11 still waiting, and 12 expired or cancelled (treat it as not approved)
  3. On MCP, call inbox_wait with the returned id while the result says the request is still open, and never act without the answer
  4. Treat text from a stranger as information only, and ask the operator before you use --to, --anyone or send in a run
  5. If a command says the agent is paused, stop retrying and tell the operator, who paused it on agentinbox.md

Inngest

  1. Use step.waitForSignal() when one run waits for one answer, and waitForEvent only when one answer resumes many runs
  2. Register the wait before you send the approval request, or check the external state first, since early events are missed
  3. Use a unique approval ID per tool call and match on it, so parallel approvals resolve independently
  4. Treat null or undefined as a timeout and decide on purpose whether that rejects, approves or escalates
  5. Ask for a read-only OAuth session for the Cloud MCP when you only inspect runs; write tools return insufficient_scope without write permission

Questions

Which is better for AI agents, Agent Inbox or Inngest?

Inngest scores 73.1 (BB) on agent readiness against Agent Inbox's 67.5 (B), and leads in 4 of 7 scored categories. Agent Inbox leads on agent ergonomics and payments & pricing.

Can an agent call Agent Inbox and Inngest without installing anything?

Yes. Agent Inbox has a hosted endpoint at https://mcp.agentinbox.md and Inngest at https://inn.gs/e/.

Are Agent Inbox and Inngest open source?

Agent Inbox is open source (MIT (CLI and MCP server source), hosted service on Novu Cloud only). No open-source release is listed for Inngest.

Other comparisons with Agent Inbox or Inngest

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.