ConvergeQA

This page confirms metadata for a ConvergeQA multi-model AI quality assurance workflow. It does not disclose submitted content or certify that the reviewed content is true.

✅ Verified: This report matches ConvergeQA's generated review record

Report Details

Report IDCQA-2026-00482

DocumentConsulting Agreement

Review Date2026-06-01 14:22 UTC

ModeIterate · converged · 1 round

PresetExample

Credit cost3 credits

Customer billing is recorded in credits; provider spend metadata is not shown on public verification pages.

Models Involved

Provider	Model	Role
Anthropic	Claude	Panel reviewer; synthesis Claude Opus
OpenAI	GPT	Panel reviewer
Google	Gemini	Panel reviewer
xAI	Grok	Panel reviewer
Perplexity	Perplexity	Panel reviewer

Review Configuration ▸

Findings by Severity (10 total)

Critical

Major

Minor

Info

Verification Metadata

Input Hash (SHA-256)

placeholder-fingerprint-a1b2c3...e9f0

Output Hash (SHA-256)

placeholder-fingerprint-a1b2c3...e9f0

Server-generated authentication code

not present on example records

What this page verifies ▸

This public page shows metadata only. It does not reveal the submitted document, private prompt text, model outputs, or account identity, and it does not certify that the reviewed content is true.

1. Canonical form. For reports created on or after April 20, 2026 (platform version v80), the input and output hashes were recorded from raw UTF-8 bytes exactly as submitted, with no whitespace collapsing, no newline normalization. Earlier reports used a whitespace-collapsed canonical form; those records remain green because the stored hashes were computed against that earlier rule.

2. Reproduce the input and output hashes. Given the exact bytes of the input document and the exact bytes of the review output, compute SHA-256 in your tool of choice:

sha256sum input.txt    # expect: placeholder-fingerprint-a1b2c3...e9f0
sha256sum output.txt   # expect: placeholder-fingerprint-a1b2c3...e9f0

Python equivalent: hashlib.sha256(open('input.txt','rb').read()).hexdigest().

3. Server authentication field. The server authentication code is computed from the stored review-record fields:

message   = input_hash + ":" + output_hash + ":" + report_id + ":" + end_time
auth_code = server_authenticator(signing_key, message.encode("utf-8")).hexdigest()

The four fields are concatenated with literal colons, no spaces. end_time is the ISO-8601 timestamp stored on the record. The authentication key is held only by ConvergeQA servers.

If you have the original source files, you can compare locally computed hashes to the metadata shown above. If the hashes differ, the local files do not match the recorded review metadata. For questions, contact [email protected] with the Report ID.

Cite This Report

Short: for social media, YouTube descriptions, footnotes

Reviewed via ConvergeQA multi-model AI review (Iterate · converged · 1 round mode, 5 model-based review passes). Audit Report CQA-2026-00482, June 01, 2026. Verify: https://convergeqa.net/verify/CQA-2026-00482

Full: for documents, blog posts, formal references

This content was reviewed using ConvergeQA, a multi-model AI quality assurance platform. The review was conducted in Iterate · converged · 1 round mode using 5 model-based review passes (Anthropic, OpenAI, Google, xAI, Perplexity). The review record, including findings, change log, and verification metadata, is available at https://convergeqa.net/verify/CQA-2026-00482 (Report ID: CQA-2026-00482, June 01, 2026).