agentleFS
odinfreeodinfree/storm-ecdsa-harness/skills/quantum-mbundo-lifetimes/SKILL.md

quantum-mbundo-lifetimes

Skill4 starsChanged 21 days agoMITView on GitHub

Assess source-proven ancilla last-use intervals for measurement uncomputation. Distinguish parked stage-level opportunities from unassessed whole-division changes and require correction-inclusive economics.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the shared contract before a trial.

    line 12
  2. Distinctive:

    Stop or defer if required gates/measurements are unsupported, dependencies cannot be proven, correction costs erase the gain, or relative …

    line 31

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Iitagsflüge (quantum-mbundo-lifetimes)

Quantum Lifetime Uncomputation

Change the cleanup architecture, not simply the implementation of a reverse gate. Read MBUndo.

Read the shared contract before a trial. Start with the five-line post; the shared native-test order, stop rules and promotion gate apply.

Stage-level negative and whole-division question

The inspected stage endings are parked: later stages still consume their values, so the proposed early measurement is ineligible. The dated estimate has partial audit coverage; it is not a measured gain or a bound on all divisions.

For a bounded reopening, freeze the parent and enumerate actual last-use suffixes. Price eligible corrections and pass native reachable-state tests before one diagnostic emission; if no useful suffix survives, stop without emitting. A diagnostic emission cannot pass the two-build promotion gate.

Recover the whole-division owner and current verdict before reopening. Analyze the interface and downstream consumers; do not sum local stage opportunities into a whole-division bound. Phase repair need not physically reconstruct every measured bit, but that observation alone supplies no cheaper circuit.

Practical lane

Inventory the two divisions and other expensive cleanup regions separately. Identify logical ancilla lifetimes even when they share a physical wire. Record producer, final quantum use, subsequent writes, dependencies, reset and reuse. Apply the paper's unitary/zero-endpoint premises to the actual region; existing measurement/feed-forward needs a separate argument. Silent allocator frees may hide endpoints from reset-based passes, but neither a free call nor sampled zeros proves eligibility.

Start with one lifetime whose cleanup dominates cost. Compare original unitary cleanup, endpoint MBU and interval MBU. Preserve the timing of corrections relative to changing controls and keep measurement results live until their final conditional use. Never measure a borrowed dirty register under a clean-ancilla rule.

First experiment: enumerate a small circuit's input and measurement branches against an independently specified channel. Include an interleaved case where another ancilla still reads the proposed measurement target. This should reject measurement before its last actual use. Then measure the full-width emitted replacement using the board's gate accounting.

Structural extension

Investigate whether an earlier valid cleanup boundary frees a peak allocation for another consumer. Combine with borrowing only after recomputing the entire lifetime graph. Record the causal change in storage or correction structure, not merely a reduced local gate count.

Stop or defer if required gates/measurements are unsupported, dependencies cannot be proven, correction costs erase the gain, or relative phase/scratch checks fail. T-gate savings in the paper are not executed-Toffoli savings on this board.

Deliver lifetime analysis, rewritten region, correction dependencies, independent channel evidence and complete-path resources.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.