Recordings, bookmarks & replay

Keep the evidence from a rare failure after the live session ends.

Reopen the capture, not the incident

Save the session as a .jproxy recording so another investigation can start from the same captured requests. Retained fields depend on capture and save modes: a stored byte count does not mean the complete payload was preserved. Check omission and truncation notices.

Bookmark a request worth repeating

  • Flag critical requests as Bookmarks during an active session.
  • Inspect saved snapshot headers and bodies, refresh saved information from newer captures, copy as curl, or replay requests within granted authority.

Choose an export for the next investigator

  • Export to HAR (HTTP Archive): Export for compatible network-analysis tools. Verify required fields and body availability for the receiving workflow.
  • Clean JSON Export: Structured summaries ready for piping into jq or custom automation scripts.

Try it

Bookmark a read-only request from a synthetic test session. Save the recording, reopen it and compare the retained status, headers and available body. Replaying the request makes a new observation; it does not reconstruct the old server state.

Choose fidelity for the question

  • Full Recording: use for a deep investigation; per-body limits still apply.
  • Metadata Only: retain structure, status and timing without bodies. Headers can still contain sensitive information.
  • Text Bodies Only: retain readable payloads such as JSON and manifests without binary media.
  • Media Focused: preserve streaming context and bounded media bodies; do not treat the file as a playable copy of the event.

Live retention settings determine what survives in the working capture; save fidelity determines what is written from that capture. Before sending an archive, reopen it and inspect the exact session the recipient will need.

FAQ: why did a replay differ from the bookmark?

Replay creates a new request through the running proxy. A server, fixture, cookie or expiring credential may have changed. Use only a read-only request in a synthetic test. Compare the new captured response with the bookmark; do not blindly replay a state-changing operation. To make a fresh capture, set up the device. For an interrupted run, see previous-session recovery.

Finish

Disable any test rewrite or replay-related effect. Review the recording’s contents before sharing; a compressed native file is not anonymized or encrypted by virtue of compression.