Skip to main content
Manusで任意のスキルを実行
ワンクリックで
GitHub リポジトリ

gbllm

gbllm には bkase から収集した 8 個の skills があり、リポジトリ単位の職業カバレッジとサイト内 skill 詳細ページを表示します。

収集済み skills
8
Stars
2
更新
2026-05-04
Forks
1
職業カバレッジ
3 件の職業カテゴリ · 100% 分類済み
リポジトリエクスプローラー

このリポジトリの skills

gbf-debug-usage
ソフトウェア開発者

Drive the gbf-debug agent CLI to inspect, step, and script Game Boy ROM execution. Use when working with .gbsess files, gbf-debug, the gb JS object, breakpoints, watchpoints, symbols, or Game Boy ROM debugging in this workspace.

2026-05-04
qat-bead-closure
ソフトウェア品質保証アナリスト・テスター

Use when implementing, reviewing, or closing gbllm QAT beads; enforces artifact-contract, differentiable Burn path, proof-test, acceptance-criterion movement, claim-to-gate, and no-future-variant checks.

2026-04-28
logging-bead-closure
ソフトウェア開発者

Use when implementing, reviewing, or closing gbllm structured logging beads, including tracing event schemas, training/export/preflight logs, subscriber capture tests, and logging adoption claims.

2026-04-27
model-contract-bead-closure
データサイエンティスト

Use when implementing, reviewing, or closing gbllm model topology or scalar gbf-model semantics beads, especially config validation, topology choices, parameter sharing, and non-Burn model behavior.

2026-04-27
asm-bead-closure
ソフトウェア開発者

Use when implementing, reviewing, or closing gbllm ASM/ISA beads in gbf-asm, including instruction modeling, sections, symbols, builders, layout, relaxation, effects, privileges, encoders, cycle models, and listings.

2026-04-27
fixture-bead-closure
ソフトウェア品質保証アナリスト・テスター

Use when implementing, reviewing, or closing gbllm fixture beads, especially tiny model/artifact/workload fixtures, deterministic helper data, and artifact assertion helpers.

2026-04-27
sequence-state-bead-closure
ソフトウェア開発者

Use when implementing, reviewing, or closing gbllm sequence-state beads, including LinearState, BoundedKv, SequenceBlock traits, exported sequence facts, byte-backed state, and recurrent state tests.

2026-04-27
feedback-harness-reflection
ソフトウェア開発者

Reflect on recent user feedback, review findings, transcript mistakes, bead closure problems, or repeated agent issues to extract reusable project guides and automated sensors. Use when the user asks to turn feedback into guardrails, harness checks, skills, AGENTS instructions, tests, pre-commit/CI checks, or next-time prevention mechanisms.

2026-04-26