HAR to Markdown
Turns a captured .har file into a readable Markdown writeup, a summary table of every request,
then one section per request with its headers in a code block. Paste straight into a PR description, bug
report, or Slack message.
About This Tool
Reuses the same HAR parsing as HAR File Viewer, then formats the result as plain
Markdown: a ## API Calls Summary table (method, path, status, time) followed by one
### section per request with its request and response headers, and request body, when the
capture included one; each in its own fenced code block.
Useful when a HAR file needs to be shared somewhere that doesn't render JSON nicely: a GitHub PR description, a bug tracker comment, or a Slack thread. For the same data as a browsable table instead, use HAR File Viewer; for an OpenAPI skeleton, see HAR to OpenAPI.
Key Features
- Summary table first: method, path, status, and time for every request at a glance.
- Per-request sections: numbered
###headings with full request/response headers. - Request body included: shown in its own fenced block when the HAR captured
postData.text. - Query strings kept in the path column: nothing is stripped, unlike the OpenAPI converter.
- Upload or paste: drop a
.harfile or paste JSON directly.
Example
A HAR with one request produces:
## API Calls Summary | Method | Path | Status | Time (ms) | | --- | --- | --- | --- | | GET | /users?page=1 | 200 | 182 | ### 1. GET /users?page=1, 200 Full URL: `https://api.example.com/users?page=1` **Request headers:** ``` Accept: application/json ``` **Response headers:** ``` Content-Type: application/json ```
FAQ
- Does it include response bodies?
- No, only the request body (when captured). Response bodies can be large and binary; use HAR File Viewer to inspect those in full.
- Is the path column the same as in HAR to OpenAPI?
- No, here the query string is kept as part of the path for full context; HAR to OpenAPI strips it to group requests by endpoint.
- Can I paste this straight into a GitHub PR description?
- Yes, that's exactly the intended use, GitHub, GitLab, and most issue trackers render the table and code blocks as-is.