Firecrawl Monitoring

Change-detection for AI agents. Enter a URL, describe what to track in plain English, and /monitor pings your agent by webhook the moment a page changes, ingesting only the diff. Source: X/@firecrawl, 2026-05-28

Monitoring by Firecrawl turns "watch this page for me" into an agent primitive. You set a goal in natural language, for example "alert me when the Claude Code docs add new slash commands," and the service tracks specific elements or whole pages with content filters and scrape behavior you control. The headline efficiency claim: up to 90% fewer LLM tokens by only feeding the agent what actually changed instead of re-ingesting the full page. Source: X/@firecrawl thread, 2026-05-28

How It Works

• Initialize via API, CLI, MCP, or dashboard.
• Schedule every 5 minutes, hourly, daily, or a custom cadence.
• Notify by webhook or email, only when watched pages actually change.
• Rich diffs: every change reports what was added, removed, and changed, plus a shareable permalink.
• Result states: page checks resolve to same, new, changed, removed, or error.
• Diff modes: markdown/git-diff, structured JSON extraction, or mixed mode.
• Webhook events: monitor.page for each page and monitor.check.completed for the reconciled run.

Source: X/@firecrawl thread, 2026-05-28; Source: Firecrawl Monitoring docs, 2026-06-30

Current Status Snapshot

Firecrawl Monitoring is a hosted API/docs feature, not a standalone package with a public semantic version. The current docs describe /v2/monitor endpoints, scrape and crawl monitors, meaningful-change judging, JSON-mode field diffs, mixed-mode diffs, email summaries, webhook delivery, check inspection APIs, and credit-based pricing. Versioning note: record the docs URL and reviewed date instead of inventing a package version. Source: Firecrawl Monitoring docs, 2026-06-30

Use Cases

Firecrawl lists keeping agents current on research papers, developer docs, feature announcements, financial filings, government regulations, and competitor pricing. Source: X/@firecrawl thread, 2026-05-28

Where It Fits Kevin's Stack

This is the push-based complement to Kevin's pull-based research tools. Where Security and Review Skills and Security and Review Skills go fetch live signal on demand, Firecrawl Monitoring inverts the loop: it watches sources and wakes an agent only on change, the same token-economy logic as Firecrawl's existing scrape products. Natural fits include the radar automations (frontend-frontier-radar, signal-radar, seo-geo-radar) and tracking docs/changelogs that feed wiki tool pages. Wire it to a webhook endpoint behind the Remote MCP Server server to route changes into the wiki ingest pipeline.

The integration rule is to track source surfaces, not random pages: docs/changelogs, pricing, release notes, regulatory pages, benchmark pages, and competitor product docs. When a monitor fires, the agent should store the diff and source URL as an ingest artifact before updating wiki pages. That preserves provenance and prevents "the page changed" from becoming untraceable chat memory.

---

Timeline

• 2026-06-30 | Deep-reviewed the bookmark thread and official docs. Added the local media artifact, recorded Firecrawl Monitoring as a hosted feature with no public package version, and connected it to source-compile/radar workflows as a push-based source-change trigger. Source: X/@firecrawl thread, 2026-05-28; Firecrawl docs, 2026-06-30
• 2026-05-28 | Firecrawl released Monitoring (/monitor): plain-English goals, element/page tracking, 5-min-to-custom cadence, webhook/email alerts, rich diffs with permalinks, ~90% token reduction. 663 likes, 705 bookmarks. Source: X/@firecrawl, 2026-05-28