Skip to main content
XyraSinclair
GitHub クリエイタープロフィール

XyraSinclair

5 件の GitHub リポジトリにある 25 件の収集済み skills をリポジトリ単位で表示します。

収集済み skills
25
リポジトリ
5
更新
2026年8月21日
リポジトリエクスプローラー

リポジトリと代表的な skills

audit-the-oracle-coverage
ソフトウェア開発者

The dangerous case is not "no oracle" — it is a WEAK oracle you falsely believe is exact. Before trusting any test suite, metric, benchmark, eval, or proof as if it settled the question, state explicitly what it CERTIFIES and what it does NOT — the gap…

2026年7月9日
carry-the-residue-forward-across-sessions
ソフトウェア開発者

Stop re-deriving from zero every session. At the close of any non-trivial work session, extract what RESISTED resolution — anomalies, reframes that failed, objections that survived, named-but-unfilled gaps, contested judgments, the unexplained core of a…

2026年7月9日
coin-the-missing-concept
ソフトウェア開発者

When the same awkward phrase keeps recurring — "the kind of bug where...", "this class of lines that...", "that pattern we keep tripping on but have no name for" — treat the repetition as residue signaling a missing concept. Coin a precise noun, give it a…

2026年7月9日
design-the-mechanism-from-the-desired-equilibrium
ソフトウェア開発者

When the task is to set rules, incentives, or metrics — "how should we score this", "what policy makes agents tell the truth", "how do we reward the right behavior" — do not propose a rule and hope it induces the outcome. Start from the behavior you want as a…

2026年7月9日
enumerate-discriminating-alternatives
ソフトウェア開発者

When one narrative starts dominating early — a bug explanation, a market story, a user motive, "the obvious cause" — stop converging and force at least four live alternatives, each still consistent with what has been seen and each carrying a prediction that…

2026年7月9日
localize-then-glue
ソフトウェア開発者

When a global attack stalls — the system is too coupled, the proof too wide, the rollout too distributed, the synthesis too brittle — stop pretending the whole thing must yield in one piece. Solve it on tractable patches, then compute exactly what fails on…

2026年7月9日
preserve-the-target
ソフトウェア開発者

The universal pre-gate that wraps every other skill. Every cognitive move TRANSFORMS the problem — tests it, reframes it, localizes it, names it, formalizes it, dimensionalizes it, persists it — and the sharpest failure is not that the move is non-executable,…

2026年7月9日
prove-the-coverage-denominator
ソフトウェア開発者

Whenever you are tempted to say "done", "complete", "fully reviewed", "all sources covered", or "the migration is comprehensive", force the denominator into the open first. State the full space of cases, sources, branches, or entities under discussion, then…

2026年7月9日
収集済み skill 14 件中 8 件を表示しています。
forward
その他コンピュータ職

Run one Drive Forward cycle — orient on the project's intent, interview if needed, hunt for issues with a statistical stopping rule, make safe well-scoped improvements, verify, and appraise worth. Use when the user says "/forward", "drive this forward", "push…

2026年7月9日
hazard-hunt
ソフトウェア品質保証アナリスト・テスター

Hunt a codebase for issues (bugs, duplication, drift, inefficiency) under a statistical stopping rule — issues beget issues, and an area is only released after enough clean spend that the upper confidence bound on the remaining rate drops below its value…

2026年7月9日
intent-ledger
その他コンピュータ職

Capture and compile a project's strategic intent from the stream of prompts, corrections, and decisions over time — deduplicating, weighting recent statements over old, and recording supersessions instead of silently forgetting. Use when bootstrapping…

2026年7月9日
worth-ledger
その他コンピュータ職

Measure whether shipped changes were actually liked and what they were worth — elicit valuations from the user, record worth against tokens spent, and feed the results back into stopping thresholds and direction weights. Use at the end of a cycle, after a…

2026年7月9日
model-sentiment
ソフトウェア開発者

Elicit what models genuinely feel and think about a codebase — unease, admiration, embarrassment — and use it as a calibrated prior for where issues live and where structure wants to go. Use at the start of a hunt, before large refactors, when consulting…

2026年7月9日
diamond-polish
ソフトウェア開発者

Raise touched code toward canonical, diamond-grade form — deduplicate, simplify isomorphically, remove efficiency losses, sharpen interfaces — guided by the von Neumann and Feynman tests. Use when improving code quality, hunting duplication, refactoring, or…

2026年7月9日
dimension-interview
ソフトウェア開発者

Identify the independent dimensions along which a project could grow, then interview the user — once, in one batch — about which directions excite them, what scope appetite they have, and what's off-limits. Use when the intent ledger is stale or silent on a…

2026年7月9日
safe-scope
ソフトウェア開発者

Discipline for making changes safe and well-scoped — blast radius estimated before editing, one concern per change, reversible checkpoints, commit early and push immediately. Use before and during any non-trivial edit, especially autonomous ones.

2026年7月9日
5 件中 5 件のリポジトリを表示
すべてのリポジトリを表示しました