---
title: "Keenable Integration"
description: "Keenable is a web search API for AI agents that runs its own crawler, index and ranking. Agents search the web with domain, publication and indexing filters, rerun a search as of a past moment, and fetch any page as clean markdown or as the answer to an extraction prompt."
url: https://flowrunner.ai/integrations/keenable
date_modified: 2026-11-03T09:29:12-08:00
---

# Keenable

[AI](https://flowrunner.ai/integrations/category/ai-llms)

Keenable is a web search API for AI agents that runs its own crawler, index and ranking. Agents search the web with domain, publication and indexing filters, rerun a search as of a past moment, and fetch any page as clean markdown or as the answer to an extraction prompt.

[Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 2 actions · API key · available

[Keenable website](https://keenable.ai/) · [Platform Documentation](https://docs.keenable.ai/) · Capability data verified 2026-11-02

1.  Weekly schedule starts the competitor pricing watch
2.  Search Web restricted to each competitor Site, Published After 7d
3.  Agent skips URLs already logged in Airtable
4.  Fetch Page with an Extraction Prompt for pricing tiers
5.  Agent compares the answer to the last recorded tiers
6.  Changed tiers posted to Slack with URL and indexing time
7.  Product marketing lead confirms before the battlecard changes

## What This Integration Enables

Keenable is for agents whose answer depends on what the web said, and when it said it. It runs its own crawler, index, and ranking rather than reselling another engine's results, and the connector exposes the two calls an agent actually needs. Search Web returns ranked results with relevance snippets, filtered to one domain, to a publication or indexing window, or rerun as of a past moment. Fetch Page turns any URL into clean markdown, or into the answer to an Extraction Prompt.

The date filters are the point. An agent can ask for pages a competitor published in the last week, or replay a search as it would have looked on the day a decision was made, which makes web-grounded work auditable after the fact. Every successful search or fetch costs one credit and failed calls are never billed. FlowRunner adds the part a search API cannot: an agent that knows when an extracted fact is too thin to act on, and stops to ask a person through [human-in-the-loop](https://flowrunner.ai/concepts/human-in-the-loop) before it changes anything downstream.

### Without FlowRunner

**Tab-by-tab checking**: Someone opens each competitor pricing page when they remember to

**Copy-paste evidence**: Findings land in a doc with no record of when the page said it

**Unknown freshness**: Nobody can tell whether a page was read today or cached months ago

### With FlowRunner

**Scheduled discovery**: Recent pages from each competitor domain surface on a fixed cadence

**Dated evidence**: Every result carries its URL, publication time, and indexing time

**Freshness on request**: The agent reads the live page when the indexed copy is not good enough

## Flows you can build with Keenable

### Weekly competitor pricing watch

Every Monday the agent runs Search Web once per competitor with Site set to their domain and Published After set to 7d. It checks each URL against the log in [Airtable](https://flowrunner.ai/integrations/airtable) with Find Record and drops the ones it has already seen. For new or changed pricing pages it calls Fetch Page with the Extraction Prompt "List every pricing tier with its monthly price and seat limit," compares the answer to the tiers recorded last time, and writes changes back with Update Record. Anything that moved goes to the product marketing lead before a battlecard is touched.

### Topic digest for the team wiki

A research agent runs Search Web on a market topic in Pro mode with Published After set to 7d, keeps the top results, and fetches each one as markdown with Fetch Page. It summarizes the set into one digest, cites every source URL with its publication time, and files it with Create Page in [Notion](https://flowrunner.ai/integrations/notion). The team reads one page instead of twenty tabs, and every claim in it points back to a dated source.

### Reproducing what the agent saw

A sales lead asks why the agent told a prospect that a competitor had dropped a free plan three weeks ago. The agent reruns the original query with Search As Of set to the moment it first answered, so pages indexed later are left out. The rerun shows exactly which results the earlier answer rested on. If the evidence was thin, the lead corrects the record with the prospect. If it holds, the answer stands with its sources attached.

## Human-in-Loop Highlight

When Fetch Page runs with an Extraction Prompt, the content it returns is only an LLM's answer about the page, not the page itself, and by default it reads Keenable's indexed copy rather than the live site. A misread price and a real price change look identical to the agent. So before a changed tier reaches the battlecard, the agent posts to the product marketing lead in [Slack](https://flowrunner.ai/integrations/slack) with Send Message To Channel: "Acme's Growth tier reads $79, down from $99, per their pricing page indexed on 28 October. Update the battlecard, or re-fetch the live page first for one more credit and compare?" The lead decides whether the evidence is good enough. The agent does the reading; the person owns what the sales team repeats to customers.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via Slack

Human decides

Agent resumes with decision

## Agent Capabilities

2 actions

### Search

1

-   **Search Web** Searches Keenable's own index and returns ranked results, each with a title, URL, a snippet highlighting the parts relevant to the query, and the page's publication and indexing times when known. Pro mode retrieves more deeply and Realtime answers faster. The agent can restrict results to one Site, limit them to a publication or indexing window using dates or relative offsets such as 7d or 3mo, and run the search as of a past moment with Search As Of so later pages are left out. Used for monitoring a domain, grounding an answer in current results, and replaying an earlier search for an audit. Each successful search costs one credit; failed calls are never billed.

### Fetch

1

-   **Fetch Page** Fetches a web page and returns its content as clean markdown, with the title, author, and publication time when known. By default it reads Keenable's indexed copy, which fails for a page Keenable has not indexed; turning on Fetch Live reads the page from its source instead. Given an Extraction Prompt, the content holds only an LLM's answer to it, for example the pricing tiers listed on the page. Used for turning a page into text for summarizing or archiving, and for pulling one fact set from a page without parsing it. Each successful fetch costs one credit.

## Frequently asked questions

### What can FlowRunner do with Keenable?

FlowRunner agents can run Search Web and Fetch Page in Keenable.

### Does connecting Keenable to FlowRunner require OAuth?

No. Keenable connects to FlowRunner with an API key, no OAuth flow required.

### Can Keenable trigger a FlowRunner workflow automatically?

Keenable doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

**Work at Keenable?** This integration exposes Keenable to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current.

## Need a custom Keenable automation? Our team can build it for you.

Tell us the process. Our implementation team at Midnight Flow builds it around how you work, tests it on your real data, and keeps it running, in your own FlowRunner workspace.

[Have us build it](https://midnightflow.ai/packages/custom?utm_source=flowrunner.ai&utm_medium=referral&utm_campaign=general&utm_content=integrations.aside&utm_term=keenable)

---
Markdown version of https://flowrunner.ai/integrations/keenable. Site index: https://flowrunner.ai/llms.txt
