# Quantum Syndrome Uncompute

> Explore syndrome-based deterministic or post-selected uncomputation when garbage support has exploitable structure; verify acceptance semantics before board use.

- Skill: `odinfree/quantum-syndrome-uncompute` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx get-skill add odinfree/quantum-syndrome-uncompute`
- Raw SKILL.md: https://api.skillmd.com/api/skills/odinfree/quantum-syndrome-uncompute/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: odinfree (https://skillmd.com/u/odinfree)
- Updated: 2026-11-16
- Page: https://skillmd.com/skills/odinfree/quantum-syndrome-uncompute

---


# Putzequipe (quantum-syndrome-uncompute)

*Quantum Syndrome Uncomputation*

Seek a smaller correction representation before proposing probabilistic cleanup. Read [the error-correction framework](references/paper.md).

Read the [shared contract](../quantum-structural-lanes/references/lane-contract.md) before a trial. Start with the five-line post; the shared native-test order, stop rules and promotion gate apply.

## Practical lane
Define the garbage function, populated support and surviving registers for one frozen boundary. Prove or exhaustively establish the relevant support at small scale. A sampled support gives a lower bound on possible affine rank; it cannot certify a small full-support rank.

Compare three routes separately: deterministic correction of every outcome, partial correction with explicit rejection, and pure post-selection. First test deterministic correction on a tiny independent channel model. Then enumerate probabilities and conditional output states for any probabilistic route, including input dependence and all rejected outcomes.

The first board gate is semantic: does the evaluator support heralded rejection, represent its probability, and charge repetitions? Ordinary incorrect outputs and post-selected failures are not interchangeable. If the rules cannot represent the proposed instrument, keep it as a research result and do not submit it as an ordinary circuit.

When repeats are meaningful, price the full failed attempt and re-preparation, not just successful correction. Unknown quantum inputs cannot generally be recreated after rejection. Do not replace the official score with T/p unless its operational assumptions and scoring contract actually justify that equation.

## Structural extension
Explore a cheaper basis of phase predicates or outcome-adapted correction with a valid fallback for every syndrome. This can offer a deterministic structural route even when post-selection is inadmissible.

Stop on unsupported rejection semantics, unproven support, destructive failure branches without valid restart, or unfavorable full-path cost. Deliver the support/rank argument, correction basis, complete instrument, probability evidence and explicit board admissibility decision.

