Free tools

Free HAR Sanitizer

Clean sensitive session details from HAR exports before sharing with engineering, support, or vendors.

Copy-paste outputs

The autonomous growth engine for your business.

Brand Armor AI finds where your customers are, decides what to create, distributes it across the channels that matter, and learns from every result.

Discover customer opportunitiesPlan campaignsCreate posts, images, and videosPublish across channelsMeasure what worksAI search visibilityDiscover customer opportunitiesPlan campaignsCreate posts, images, and videosPublish across channelsMeasure what worksAI search visibility
Social publishingBlog and CMS deliveryEmail outreachAutopilot workflowsBrand IdentityWeekly reportsSocial publishingBlog and CMS deliveryEmail outreachAutopilot workflowsBrand IdentityWeekly reports
Shopping IntelligenceCompetitor AnalysisBrand IdentityShopping IntelligenceCompetitor AnalysisBrand Identity

Tool 01

HAR Sanitizer

Redact cookies, auth headers, tokens, and sensitive parameters before sharing HAR files.

HAR Sanitizer
Remove auth headers, cookies, tokens, and sensitive query data from HAR payloads before sharing.
Sanitized output
// Sanitized HAR output will appear here

How it works

HAR Sanitizer: methodology and worked example

How this tool computes its result

Recursively walks a parsed HAR JSON structure (arrays and objects) and, for any object key whose lowercased name contains a sensitive substring — authorization, cookie, set-cookie, token, apikey/api-key/x-api-key, password, secret, session, or bearer — replaces that key's entire value with the literal string '[REDACTED]', incrementing a redaction counter. As a special case, any key literally named 'url' has its value parsed as a URL, and each query-string parameter whose key OR value matches the same sensitive keyword list is rewritten to '[REDACTED]' in place, with the reconstructed URL string substituted back in.

Worked example

Given a HAR fragment with a header object {"name":"Authorization","value":"Bearer abc123"} and a request url of 'https://api.example.com/login?token=xyz789&user=jane', the 'name' key's value 'Authorization' is not itself redacted (the check runs on key names, not values), but a header keyed literally 'authorization' elsewhere would be. Meanwhile the token=xyz789 query parameter is redacted to token=[REDACTED] because 'token' matches the keyword list, and the redacted-fields counter increments accordingly.

When not to use this tool

The input must be valid JSON parseable by JSON.parse — a malformed or truncated HAR export (a common failure mode when copy-pasting very large HAR files from browser DevTools) fails outright with a generic 'HAR must be valid JSON' message rather than attempting any partial or best-effort sanitization.

Common mistakes

  • - Assuming header VALUES are scanned for secret-looking content — the redaction logic only inspects field NAMES (like 'authorization' or 'cookie'); a header named something generic like 'X-Custom' whose value happens to be a real secret is never caught.
  • - Pasting only a partial request/response fragment instead of a full HAR log — the sanitizer walks whatever JSON shape it's given without requiring a log.entries structure, but downstream tools expecting a complete HAR file may reject the sanitized fragment.
  • - Expecting response bodies to be scanned for embedded secrets — only object keys and URL query parameters are inspected; base64-encoded content or raw response body text is passed through untouched.

Ready to put your growth work on autopilot?

Connect your business once, then let Brand Armor AI find opportunities, create content, publish it, and learn from what brings customers in.