# HubSpot agent tool config describes the run you can prove

> Write the agent tool's type, action description, inputs, and outputs as a contract you can demonstrate on every successful run.

- Canonical HTML: https://growth.iangoh.com/growth-ideas/hubspot-agent-tool-config-describes-the-run-you-can-prove/
- Source: [developers.hubspot.com](https://developers.hubspot.com/docs/apps/developer-platform/list-apps/agent-tool-listing-requirements)
- GrowthDex source hub: [HubSpot Docs: Agent tool listing requirements](/sources/hubspot-docs-agent-tool-listing-requirements-developers-hubspot-com/)
- Last checked: 2026-06-05T06:15:00Z
- Rarity: rare
- Budget: free
- Channels: Marketplaces, Product-led Growth, Engineering
- Stages: tool schema, output contract, uat, agent reliability


## Why this can grow

The weakest agent listings treat configuration as metadata and the actual run as a separate concern. HubSpot treats them as one thing. `toolType` has to match the operation, `llmConfig.actionDescription` has to follow description guidance, and every configured output has to be returned in the promised format. That forces the product page, tool schema, and live behavior to agree. When they agree, the review process is faster and the buyer sees a workflow instead of a prompt-shaped guess.

## Ian's take

From scaling consumer platforms across MENA and Southeast Asia, my default is to distrust growth work that only looks good in a slide. My bias is to treat this as a small market test first. Make the audience narrow, make the promise concrete, and let the first real response decide whether it deserves more work. I would run it small enough to learn quickly, then only scale the parts that real users repeat, save, reply to, or buy from. For this tactic, I would watch one clear growth signal before putting more time or budget behind it.

## Action plan

1. Define one narrow startup segment where hubspot agent tool config describes the run you can prove can create a measurable lift.
2. Turn the tactic into one offer, page, campaign, or workflow for the Marketplaces and Product-led Growth channel.
3. Use the evidence from developers.hubspot.com to set the first version of the message, format, and audience.
4. Launch a small test for 7 to 14 days with one success metric: one measurable growth signal.
5. Review the result, keep the winning message, remove weak variants, and turn the learning into a repeatable growth playbook.

## Source-backed example

HubSpot says `GET_DATA`, `GENERATE`, and `TAKE_ACTION` should match the real job of the tool, and if the output config includes a `success_message` string the tool should return that property in the expected format on every run.

## Adjacent tactics in the same lane

- [HubSpot agent tool scope only for the context you use](/growth-ideas/hubspot-agent-tool-scope-only-for-the-context-you-use/) - same source, 2 shared channels
- [HubSpot agent tool front-office use case before clever demo](/growth-ideas/hubspot-agent-tool-front-office-use-case-before-clever-demo/) - same source, 1 shared channel
- [HubSpot agent tool names the action, not your company](/growth-ideas/hubspot-agent-tool-name-the-action-not-your-company/) - same source, 1 shared channel
- [HubSpot agent tool three-minute review video before approval](/growth-ideas/hubspot-agent-tool-three-minute-review-video-before-approval/) - same source, 1 shared channel

## Read GrowthDex essays

Browse the plain-English essay index at [GrowthDex Blog](/blog/).

## Related GrowthDex essays

- [The agent tool listing should survive the approval screen](/blog/the-agent-tool-listing-should-survive-the-approval-screen/) - ai products, marketplaces, brand trust

## Advisory

If you want help turning this into a working growth system, Ian Goh offers advisory at https://iangoh.com/advisory.