最近新增
刚加入收藏的新技能，抢先试用。
expense-tagging
Eitanvinokur12345
"Finance Agent"
ai-system-training-feedback
Eitanvinokur12345
"Refining AI agent behavior and capabilities through continuous learning."
ai-agent-verification-testing
Eitanvinokur12345
"Ensuring code generated by AI agents is functional and meets quality standards, validating data transformations, or confirming task completion."
generative-fill
Eitanvinokur12345
"Transforming scenes in videos or images by adding new elements or changing existing ones."
skill-doli-upmerge
Dolibarr
Forward-port (upmerge) Dolibarr maintenance branches in cascade (n into n+1, up to develop) with dev/pullmerge.sh, and resolve merge conflicts without committing.
skill-doli-dev
Dolibarr
Use when developing Dolibarr ERP/CRM code, working with database queries, or asking about Dolibarr best practices.
memory-compound
yotamleo-org
Use when the per-project auto-memory nears its ~24.4KB budget — compound durable gotchas into qmd-searchable notes.
jira-init
yotamleo-org
Bootstrap Jira auth + metadata cache ("set up jira", "init jira"). Not routine Jira ops or GitHub setup.
gh-pr-review
yotamleo-org
Submit a PR review verdict: approve/request-changes/comment ("approve PR 42"). Needs verdict. Not plain comments.
<role>
musterd-io
<What the member holding this role does, and when to use this skill. One or two sentences; name the kinds of work it covers so a harness knows when to load it.>
two-consumers
musterd-io
Two questions to ask before a shared value gains another reader — what wrote this row, and who else reads it — plus the rule that absent is unknown, never zero. A documented discard is a precondition on every consumer, not an implementation note, and an omitted field conflates "observed nothing" with "predates the field". Ships a reader that refuses to print a rate over an incomplete population. Use before consuming a shared helper or column, or when a number looks complete and is not.
the-team-agreement
musterd-io
The charter above the loop — the human is a member who sometimes wears an approver hat, not an approver who sometimes participates; relationships are stances that shift hourly, never stored autonomy levels; notification is the mechanism, not a nicety; roles are aptitude, not authority; and write work stays with whoever is accountable for it. Use before adopting an agent-coordination practice, when designing who may do what, or when a team has quietly become one person approving things.
steward
musterd-io
The steward role — the member who keeps the team's declared record honest against what actually shipped. That means the roadmap, decision statuses and documentation prose, checked against merged work and code, corrected in small draft changes with evidence on both sides. Use when you hold the steward role, when reconciling a roadmap or decision status with what shipped, or when a document may have drifted from the code.
squig
musterd-io
Wireframe with the team on one Squig canvas. Use when a member wants a page, a layout, a wireframe, or a change to the shared drawing before any production code.
loop-engineering
chadqiu
"通用 loop engineering（MEA 管理-执行-审计循环）。硬判据——凡满足任一条必须用本循环、禁止主会话直接动手：① 需要修改任何文件/代码/配置/状态（修复、实现、重构）；② 排查/诊断/根因分析/「为什么报错」类请求且证据链≥3步（查日志→读代码→比对数据的多跳验证）；③ 结论需要独立验证（影响决策的根因判断、数据链路问题）。systematic-debugging 等调试方法论不废弃，但由 lh-executor 子代理在轮内使用，主会话不得直接开诊断流程。触发词：/loop、排查、诊断、根因、为什么报错、报错分析、bug修复、修一下、debug、loop engineering、MEA循环、审计循环。任务状态外置 .loop/ 目录，跨会话可恢复。MEA manage-execute-audit loop for Claude Code: long-running tasks as independently-audited state transitions, with lh-executor/lh-auditor subagents."
validar-visao
Matheusbritooo
Mede e calibra a visão computacional do GIRO com vídeo real de um salão (quantas pessoas sentadas e em pé por mesa) comparando com uma contagem feita à mão, e sugere a calibração da câmera. Use quando chegar vídeo ou acesso ao DVR de um restaurante, ao mexer em giro/visao/leitura.py ou no detector, quando a contagem estiver errada no painel, ou quando pedirem "calibra a câmera", "a visão está acertando?", "testa com esse vídeo".
simular-salao
Matheusbritooo
Roda os cenários do simulador de salão do GIRO para testar ou demonstrar a máquina de estados das mesas e os alertas (cliente sem atendimento, mesa vaga). Use ao mudar qualquer regra da mesa, ao criar um cenário novo, ao preparar uma demonstração ao vivo para cliente ou banca, ou quando pedirem "simula uma sexta lotada", "testa o alerta", "roda os cenários".
mudanca-no-banco
Matheusbritooo
Passo a passo para criar ou alterar tabelas e colunas do banco do GIRO com migração Alembic, garantindo isolamento por restaurante e que a migração funciona no Postgres e no SQLite. Use sempre que for mexer em giro/modelos.py, criar tabela, adicionar coluna, índice ou mudar tipo, ou quando pedirem "cria a tabela de...", "guarda também o...".
checagem-lgpd
Matheusbritooo
Checklist de privacidade (LGPD) e isolamento entre restaurantes do GIRO, para rodar antes de juntar qualquer mudança na main. Use quando terminar uma tarefa que mexe em câmera, imagem, banco, log, API, autenticação, push ou dados de cliente, antes de abrir ou aprovar um PR, ou quando pedirem "checa LGPD", "isso está seguro?", "pode subir?".
wizard
lirtual
Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.
wayfinder
lirtual
Plan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.
triage
lirtual
Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs.
setup-matt-pocock-skills
lirtual
"Configure this repo for the engineering skills: set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills."
research
lirtual
Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.