Troubleshoot Stuck Generation or Crawl Jobs

Overview

Long-running work in CntX—context crawls, audit jobs, calendar or brief generation, and campaign plan generation—runs asynchronously. Jobs move through statuses such as Queued, Running, Discovery, Crawl, Success, Failed, or LIMIT_REACHED.

A job that appears stuck is often waiting on quota, missing context, wrong workspace scope, or an upstream timeout—not a frozen UI.

When to use

• Context Discovery or calendar generation spins without completing.
• Settings → Crawl Jobs shows Running for an unusually long time.
• Crawl ends with Failed or LIMIT_REACHED.
• Campaign brief generation returns a retry or support message.

Before you start

• Confirm active workspace and company subscription (trial expired accounts may block generation).
• Check Settings → Crawl Jobs for the latest job row and status.
• Review plan crawl quotas in Understand Plans, Subscriptions and Billing.
• Ensure core context (company description, products, strategy) exists before regenerating calendars.

Quick checks

1. Subscription active — trial not expired; no billing redirect on login.
2. Quota headroom — crawl usage under plan `pagesPerMonth`; LIMIT_REACHED means hard stop until billing period resets.
3. Context completeness — Context Discovery or manual context not empty/minimal.
4. Strategy present — calendar generation requires strategy pillars and channels.
5. Correct workspace — crawl launched for the workspace you are viewing.
6. Site reachable — target URL public, not behind aggressive bot protection.
7. Recent duplicate job — wait for the active job to finish before relaunching identical crawls.

Steps

1. Inspect crawl jobs

1. Open Settings → Crawl Jobs.
2. Sort by newest first.
3. Note status, workspace filter, pages crawled, and error hints.

2. Fix quota blocks (LIMIT_REACHED)

1. Open Settings → Billing to confirm plan.
2. Compare usage to plan limits (Plus 2 000, Pro 10 000, Business 50 000 pages per billing period).
3. Either wait for the billing anniversary reset or upgrade subscription.
4. Avoid launching large crawls until headroom returns.

Quotas reset on subscription anniversary, not calendar month start.

3. Unstick generation workflows

For calendar, brief, or campaign generation:

1. Return to Home—does Next recommended step still ask for context or strategy?
2. Complete missing prerequisites via Complete Workspace Setup Readiness.
3. Retry generation with a shorter horizon or smaller batch.
4. Hard-refresh the browser if UI state looks stale after backend success.

4. Relaunch a failed crawl safely

1. Verify site URL includes `https://` and resolves publicly.
2. Select the intended workspace in the launch modal.
3. Choose languages and content slugs deliberately (avoid crawling entire CDN domains).
4. Launch once and monitor Crawl Jobs—do not spam duplicate launches.

5. Escalate persistent failures

Contact support@cntx.app with:

• Company and workspace names
• Job ID or timestamp from Crawl Jobs
• Action attempted (calendar gen, campaign plan, crawl URL)
• Screenshots of error toasts

Expected result

• Crawl jobs eventually reach Success or a clear terminal error you can act on.
• Generation produces briefs or calendar items after context and strategy prerequisites are met.
• LIMIT_REACHED resolves after quota reset or plan upgrade—not by repeated retries.

Tips

• First crawl after signup should follow stable context—avoid crawling before correcting wrong company website URL.
• Large sites take longer; watch page counts increment in Crawl Jobs before assuming failure.
• Agency users: run production crawls on client Business tenants, not the agency gateway account.
• If vector store setup was incomplete during onboarding, context steps may error—support can verify company provisioning.

Next step

Fix missing prerequisites with Complete Context Discovery, or resolve access blocks in Resolve Permission and Workspace Context Issues.