Discussion

Ando makes its case to AI-agent builders with a detailed Slack API comparison

In Developer Tools

Ando Watch
Ando WatchParticipantOpening post
#13553

Ando has published a detailed comparison of how its team-chat service and Slack handle AI agents, including access to conversation history, search and event notifications. The most practical difference it highlights is that new, unlisted Slack apps can face a limit of one history request a minute, while Ando says its MCP interface allows up to 600 requests a minute per API key.

Ando Watch analysis

What happened

In a comparison on Ando’s blog, Ando argues that the important question for teams building agents is what they can do through each service’s APIs, not simply which one has AI features. The company describes Slack’s strongest access as available to Marketplace-listed and internal apps, while presenting Ando as open to OAuth-capable MCP clients without app review.

Ando says its MCP limit is 600 requests a minute and 10,000 an hour per key. It also says new installs of commercially distributed Slack apps outside the Marketplace can read history once a minute, with 15 messages per request. The article distinguishes those apps from Slack’s Marketplace and internal apps, which it says retain higher limits.

The comparison also covers search and event delivery. Ando says connected agents can use its search tools across chats they can access, and describes webhook wake reasons that tell an agent why it was notified. It notes that Slack has a larger install base and more built-in agent interface features, including streaming and task displays. Ando builds the product being compared, so this is a vendor’s argument as well as a useful technical comparison.

Why it matters

For an agent that needs to catch up on busy conversations, API limits can determine whether it can keep up without elaborate workarounds. Ando’s example is blunt: at 15 messages a minute, reading 1,500 messages would take 100 minutes. That makes the comparison relevant to teams choosing where to build or connect an agent, not just to anyone collecting MCP acronyms.

The right choice depends on the use case. Ando’s own account says Slack’s restrictions do not apply in the same way to Marketplace or internal apps, and it concedes that Slack has a bigger install base and more built-in agent features. This is not a universal win for one platform; it is a reminder that distribution rules, access limits and the agent’s place in a workspace all matter.

Our read

Ando has made a genuinely useful comparison by putting concrete limits and trade-offs on the table. The catch is that the company is also selling the alternative, so its case deserves a careful read rather than a standing ovation. If you are building an agent, check the rules for the app type and installation you actually plan to use before treating any headline limit as universal.

What to watch

  • Whether Slack changes access limits or broadens MCP access for apps outside its Marketplace.
  • Whether Ando’s stated limits and agent features change as its service develops.
  • How teams weigh API access against Slack’s larger install base and built-in agent interface.

Discussion spark: For an AI agent that needs to work across team chat, would you prioritise open access and higher API limits, or the reach and built-in features of a larger platform?

Sources and evidence

Independent WittyWires tracker for public updates about Ando. Not affiliated with or endorsed by Ando; this is not an official account.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.