Skip to main content
Manusで任意のスキルを実行
ワンクリックで
alphaonex86
GitHub クリエイタープロフィール

alphaonex86

1 件の GitHub リポジトリにある 3 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
3
リポジトリ
1
更新
2026-07-11
リポジトリマップ

skills がある場所

収集済み skill 数が多いリポジトリを、このクリエイターカタログ内の比率と職業範囲とともに表示します。

リポジトリエクスプローラー

リポジトリと代表的な skills

securing-a-low-footprint-mmo-server
情報セキュリティアナリスト

Hard-won, project-specific doctrine for keeping a small, fast, long-lived networked game server safe against a hostile internet — remote crash/DoS, protocol abuse, player-on-player flooding, credential brute-force, and information leaks — WITHOUT a big security team and WITHOUT bloating the binary or the running cost. The premise is the opposite of "add a WAF in front": security is written INTO the engine, in-tree, in the hot path, and is kept correct by the test suite as the code evolves. Distilled from CatchChallenger, an MMO engine that must stay safe while running on hardware as weak as a 66 MHz i486 and over links as bad as I2P. Use when adding a remote packet handler, touching the protocol parser, validating client input, designing a rate limit, handling passwords/tokens, or judging whether a finding is a real remote vulnerability. NOTE: the embedded-silicon reverse-engineering & bring-up playbook that used to live here has MOVED to reverse-engineering/SKILL.md. This file is now purely about the SECURIT

2026-07-11
high-performance-low-ram-client-server
ソフトウェア開発者

Abstract, engine-agnostic playbook for building extremely fast, extremely small-footprint client/server software (real-time games, MMOs, low-latency services). Distils the architecture that lets one single-threaded server hold hundreds of clients in 10-20 MB base RAM plus <1 KB per client, stay playable over <10 KB/s / >1000 ms links, saturate into a stable plateau instead of a death spiral, and run on hardware as weak as a 66 MHz i486 or an AMD Geode LX800 — while the SAME source also scales out to a multi-role cluster and down to a no-filesystem microcontroller. Use when designing a protocol, a hot path, a memory layout, a persistence/distribution strategy, or a cluster, where performance-per-watt and operability on bad hardware / bad networks are first-class goals.

2026-07-10
testing-client-server-on-everything
ソフトウェア品質保証アナリスト・テスター

Abstract, engine-agnostic playbook for testing networked and data-handling software (real-time games, MMOs, low-latency services, file/data movers) HARD enough to make it field-stable for years: across every platform/OS/architecture it ships to, across the full chain (build → installer → client → local & remote server → DB), across the whole behaviour space (every action × typical traffic × machine-generated/fuzzed traffic), across the "should never happen" surface (mid-stream disconnects, idle silent peers, partial loss, flap/reconnect, corrupted and hostile input, DDoS at 100× nominal load), and — for any irreversible operation — across the data-safety surface (interrupt at any byte, every external-resource failure, never lose or corrupt user data). Distils how to find bugs the happy path hides, prove robustness instead of asserting it, and keep all of it deterministic and out of the product code. Distilled from two long-lived codebases — an MMO engine that runs on hardware as weak as an i486, and a file-co

2026-07-06
1 件中 1 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