# Independent reproduction protocol KB-INTERNAL-DACH-20260831

**Published:** 2026-08-31  
**Target run:** `KB-INTERNAL-DACH-20260822`  
**Status:** Open protocol; no qualifying independent report received as of 2026-08-31

## Purpose

This protocol enables an unaffiliated team to test whether a second authorized operator can retrieve the same versioned definition and reproduce a first operator’s result. Publishing it is not an independent reproduction.

## Materials

- Desk log `KB-INTERNAL-DACH-20260822`
- Aggregate CSV `aggregate-KB-INTERNAL-DACH-20260822.csv`
- Verification script `verify-KB-INTERNAL-DACH-20260822.py`
- A separately obtained authorized or public dataset
- A versioned definition pack and two independently identified operators

The private sanitized composite behind the first-party run is not published.

## Procedure

1. Declare affiliations, commercial conflicts, operator roles, source grain, row count, version, retrieval date, and file hashes.
2. Publish the pack owner, permissions, definition, exclusions, version, hash, and bind mapping.
3. Have operator A run the goal and retain prompt, model version, SQL, passages, and outputs.
4. Without a private handoff, have operator B reopen the same pack and run the same goal.
5. Independently recalculate both results and test authorized as well as denied access.
6. Publish matches, conflicts, permission failures, timing, limitations, and non-sensitive artifacts.

## Acceptance criteria

A qualifying report must be authored outside InfiniSynapse, disclose conflicts, identify source and pack versions, document two distinct operator roles, expose reproducible arithmetic, and report both access outcomes. Running only the first-party CSV verification script does not qualify.

Send a public report URL or documented contradiction to [zhuhl@infinisynapse.com](mailto:zhuhl@infinisynapse.com). InfiniSynapse may link qualifying confirming or conflicting reports.