Audit Failures

This page covers common audit failures in JetStack AI. Audits involve scanning a HubSpot portal across 450+ data points, generating scores, and producing AI-powered insights. Issues can occur at any stage of this pipeline. Each entry follows a Symptom / Cause / Solution format.

Processing Issues

Stuck Processing

Symptom: The audit has been in “Processing” status for longer than expected (more than 15 minutes for Quick Mode, more than 30 minutes for Full Mode) and is not progressing.

Cause: An audit can get stuck when:

• A specific analysis block encounters an unrecoverable API error
• The HubSpot API is experiencing widespread slowness or downtime
• The portal connection was interrupted during processing (token expired mid-audit)
• A very large portal with thousands of assets causes individual blocks to take longer than expected

Solution:

1. Check the audit’s progress detail to identify which block is stalled
2. Wait an additional 5-10 minutes — some blocks process large datasets that take time
3. If the audit remains stuck, cancel it from the Activity Log and relaunch
4. Verify the portal connection is active by checking the Portals page
5. If the issue recurs on retry, try running the audit during off-peak hours

Block Failure

Symptom: The audit completes but one or more blocks show as “Failed” in the results.

Cause: Individual blocks can fail while the rest of the audit succeeds. Common causes:

• The block requires data from a HubSpot feature that is not enabled on the portal (e.g., Service Hub blocks on a portal without Service Hub)
• The HubSpot API returned an error for the specific data points in that block
• The block requires extension-captured data that was not provided
• Rate limiting caused retries to exhaust before the block completed

Solution:

1. Review the failed block details in the audit results to see the specific error
2. If the block requires a HubSpot feature not available on the portal, this is expected — exclude that block from future audits by using a custom template
3. For extension-dependent blocks, ensure you completed the extension capture step
4. Retry the audit — transient API errors often resolve on the next run

Timeout

Symptom: The audit fails with “Audit timeout” before completing all blocks.

Cause: The audit exceeded the maximum processing duration. This happens with:

• Full Mode audits on very large portals (100+ workflows, 1000+ contacts, etc.)
• Portals where the HubSpot API is consistently slow to respond
• Multiple concurrent audits running on the same portal

Solution:

1. Retry the audit — timeout thresholds are generous, and transient slowness usually resolves
2. Switch from Full Mode to Quick Mode if trend data is not critical
3. Use a custom template with fewer blocks to reduce processing time
4. Avoid running multiple audits on the same portal simultaneously

Extension Capture Issues

Extension Capture Interrupted

Symptom: The audit is stuck in “PENDING_DATA_CAPTURE” state, or the extension capture started but did not complete.

Cause: Extension data capture can be interrupted when:

• You navigate away from the HubSpot page during capture
• The browser tab is closed or refreshed during capture
• The extension loses its authenticated session
• A browser extension conflict interferes with the JetStack AI extension

Solution:

1. Open the JetStack AI extension on a HubSpot page in the correct portal
2. Navigate to the audit task in the extension sidebar
3. Click “Resume Capture” to restart the data capture from where it left off
4. If the capture fails again, try disabling other browser extensions temporarily
5. Ensure your browser is up to date and has sufficient memory

Missing Data Points

Symptom: The audit completes but certain data points show as “No data” or “Unavailable.”

Cause: Data points can be unavailable when:

• The data point requires extension capture and the extension was not used
• The HubSpot portal does not have the feature enabled (e.g., Social Media data points on a portal without Marketing Hub)
• The data point references historical data that does not exist yet (e.g., email performance metrics on a new portal with no sent emails)
• The HubSpot API does not return data for that specific endpoint

Solution:

1. Check whether the data point requires extension capture (indicated in the template configuration)
2. Verify the portal has the required HubSpot Hub and tier for each data point
3. For historical data points, ensure the portal has sufficient historical data (at least 30 days for trend metrics)
4. Unavailable data points are excluded from scoring and do not penalize the portal’s overall score

Tier and Permission Issues

Tier Mismatch

Symptom: The audit fails with “Tier mismatch” or individual blocks fail because features are not available.

Cause: The audit template includes blocks that require HubSpot features not available on the portal’s subscription tier. For example:

• Custom report analysis on a Starter portal (requires Professional)
• Workflow analysis on a Free portal (requires Professional)
• Custom object analysis on a Professional portal (requires Enterprise)

Solution:

1. Check the portal’s HubSpot tier in the portal settings
2. Use an audit template that matches the portal’s tier
3. Create a custom template that excludes blocks requiring higher-tier features
4. Pre-built templates indicate which HubSpot tier they require

Rate Limits

Symptom: Audit blocks fail intermittently with rate limit errors.

Cause: Audits make many API calls across all blocks. Combined with other JetStack AI operations or third-party integrations, this can exceed HubSpot’s 100 requests/10 seconds limit.

Solution:

1. Ensure no other JetStack AI operations (imports, deployments) are running on the same portal during the audit
2. Retry the audit — JetStack AI’s exponential backoff handles most rate limit scenarios
3. If rate limits persist, try running the audit during off-peak hours
4. For portals with heavy third-party integration traffic, coordinate audit timing with other systems

Report and Insights Issues

Report Generation Failure

Symptom: The audit completes successfully but the report fails to generate with “Report generation failure.”

Cause: Report generation can fail when:

• The audit data is incomplete (too many failed blocks)
• Custom branding assets (logos, fonts) fail to load
• The report template encounters an edge case with the data structure
• A transient server error during PDF/HTML generation

Solution:

1. Retry report generation from the audit detail page (you do not need to re-run the entire audit)
2. If you are using custom branding, temporarily switch to default branding to isolate the issue
3. Check that your brand assets (logo images) are valid and accessible
4. If the audit has many failed blocks, the report may not have enough data — re-run the audit first

AI Insights Failure

Symptom: The audit report is generated but AI insights are missing for one or more blocks, or the executive summary is empty.

Cause: AI insights generation can fail when:

• The block data is insufficient for meaningful analysis (e.g., a block with only 1-2 data points captured)
• The AI service (Claude Haiku) experienced a transient error
• The audit data contains unusual edge cases that the AI prompt does not handle
• Rate limiting on the AI service during high-demand periods

Solution:

1. Retry AI insights generation from the audit detail page
2. Check which specific blocks are missing insights — blocks with very little data may not produce meaningful analysis
3. If the executive summary is missing, retry the summary generation separately
4. If insights consistently fail for a specific block, that block’s data may need review — check the raw data in the audit details

Still Stuck?

If none of the above solutions resolve your issue:

1. Open the Activity Log and click on the failed audit task
2. Note the task ID and which specific block or phase failed
3. Contact support at team@jetstack.ai with the task ID and a description of the failure