Reqvexa.
Reqvexa explains the state behind a captured HTTP request instead of guessing how to replay it. It preserves HAR request order and raw fields, matches runtime values with temporal evidence, expands prerequisite requests recursively, and keeps unresolved client-side values visible with an explicit status.
Architecture
Small boundaries. Explicit behavior.
HarProvider loads HAR 1.2 captures without collapsing request order or duplicate fields.
Candidate extraction normalizes runtime values across headers, cookies, URLs, bodies, and selected HTML.
ProvenanceEngine links a target value to earlier responses and records the evidence behind the relationship.
Planning and reporting recursively expand prerequisites while masking sensitive values and preserving unknowns.
Implementation
The important path, in code.
A representative excerpt showing the project's approach. The complete implementation and history remain available in the public repository.
from reqvexa.capture import HarProviderfrom reqvexa.provenance import ProvenanceEnginefrom reqvexa.planning import plan_flowcapture = HarProvider().load_file("capture.har")target = capture.requests[-1]analysis = ProvenanceEngine().analyze( capture, target_request_id=target.id)plan = plan_flow(analysis, target.id)Capabilities
Preserves request order, duplicate headers, and raw HAR fields
Traces cookies, authorization, headers, query, JSON, form, and HTML state
Matches URL-decoded, Base64/base64url, and Bearer representations
Records temporal versions with evidence, confidence, and explicit statuses
Ships CLI reports and an optional Flet desktop viewer