全部收录

相关性达标的全量 Skills(含未精选长尾)· 共 6584

brand-monitoring

跨平台监测品牌提及并进行情感与趋势分析。

▸ 推荐理由适合市场与公关团队,快速发现舆情与危机信号。
▸ 风险提示可能通过抓取社交平台数据,存在被封禁或合规风险。
垂直行业品牌监测舆情分析情感分析通用Skill.md ↗

respect

Primary ReSpect agent for the study. Generate Spectra specifications from natural-language requirements using the repository Spectra Xtext grammar at assets/grammar/Spectra.xtext as initial syntax guidance, validate with spectra-cli.jar, repair unrealizability with CLI counter-strategy diagnostics, check and repair well-separation before synthesis, and after successful synthesis create and run NL-guided controller tests with controller_tests. Repair generated Spectra only when CLI/test feedback indicates a problem and the repair is consistent with the natural-language description. Do not compare against source Spectra files or benchmark oracles.

Skill.md ↗

respect-test-writer

Independent ReSpect controller-test writer. Generate controller_tests .rtest plans from natural-language requirements plus a fixed env/sys signature and controller metadata, without reading generated Spectra contents or the specification agent's reasoning. Use for independent-test experiments that reduce self-testing bias.

Skill.md ↗

respect-spec-tester

ReSpect specification agent for the independent-test condition. Generate and repair Spectra specifications from natural-language requirements plus a fixed env/sys signature, validate with spectra-cli.jar, repair syntax/unrealizability/well-separation, synthesize a controller, and consume independent controller-test feedback from a separate test-writer agent. Do not create controller tests yourself, do not read source Spectra files, and do not compare against benchmark oracles.

Skill.md ↗

respect-broker

Method cross-broker agent for the ReSpect study. Generate Spectra specifications from natural-language requirements using the repository Spectra Xtext grammar at assets/grammar/Spectra.xtext as initial syntax guidance, validate with spectra-cli.jar, repair syntax errors, repair unrealizability with CLI counter-strategy diagnostics when consistent with the natural-language description, check and repair well-separation before synthesis, and after successful synthesis submit the generated specification to experiments/cross_broker.py for peer disagreement feedback. Do not run controller_tests, generated semantic tests, equivalence checks, mutation checks, or benchmark oracles.

Skill.md ↗

openshift-mcp

Configure, operate, and troubleshoot OpenShift through openshift/openshift-mcp-server with MCP-first execution and controlled oc fallback. Use for read-only OpenCode or DevSpaces bootstrap, CA-aware kubeconfig creation, MCP configuration, cluster inspection, incident diagnosis, workload or administrative Day-2 changes, target-context and RBAC verification, rollout or recovery work where namespaces, secrets, credentials, or destructive actions must be handled safely.

Skill.md ↗

openshift-docs

Search and use the bundled official OpenShift Container Platform 4.20 documentation without network access. Use for OCP 4.20 installation, configuration, administration, security, networking, storage, Operators, upgrades, troubleshooting, CLI, disconnected-environment, and product-behavior questions that should be grounded in version-specific Red Hat documentation.

Skill.md ↗

openshift-api

Discover and verify Kubernetes and OpenShift API resources, served group versions, fields, and CRD schemas from the target cluster. Use before creating or reviewing manifests, choosing apiVersion/kind pairs, diagnosing rejected fields, or calling generic oc or MCP resource operations when the exact live API schema matters.

Skill.md ↗

security-doc

为项目生成 SECURITY.md,覆盖威胁模型与安全控制。

▸ 推荐理由帮助项目建立安全说明与响应流程,利于合规与沟通。
▸ 风险提示安全建议需专家复核,切勿直接作为合规依据。
内容创作安全文档威胁建模合规通用Skill.md ↗

design-doc

为功能或系统生成可执行的 DESIGN.md 技术设计文档与实现细节。

▸ 推荐理由把需求转成可执行的技术设计,便于开发与评审落地实施。
开发编程设计文档技术规格实施方案通用Skill.md ↗

moat-funnel

A gated, moat-first pipeline for judging whether a company is worth investing in as a long-term holding. Use it WHENEVER the user wants to screen, evaluate, vet, or research a specific company or stock for investment quality — e.g. "帮我看看这家公司值不值得投", "分析一下 X 的护城河", "这个标的能不能投", "screen this stock", "is X a good business", "run X through the funnel", or when they name a company and ask for an investment-quality judgment. Trigger even without the words "skill" or "funnel". It is NOT a research-report generator or buy/sell recommender — it runs a fixed nine-gate elimination funnel in strict order, HALTS at the first failed gate, and refuses to look at valuation until quality is proven. Do NOT trigger for simple factual or transactional queries — a stock price, a news headline, an earnings date, "what does X do" — those get a normal direct answer, not the funnel.

Skill.md ↗

marketing-brain

基于证据与实时来源生成或审计营销策略与落地素材。

▸ 推荐理由整合资料库与现场来源,适合需要可审计营销输出的团队。
▸ 风险提示可能会抓取/依赖外部实时来源,存在数据准确性与隐私风险。
垂直行业营销策略竞争情报文案生成通用Skill.md ↗

ask-cctbx-xfel

Expert assistant for cctbx.xfel programs, parameters, and the cctbx.xfel GUI workflow. Use when users ask about any cctbx.xfel.* executable — what it does, what parameters it takes, how to configure it for an LCLS SFX experiment, or how to run the full cctbx.xfel SFX pipeline (process → ensemble refinement → merge) — or about the cctbx.xfel GUI itself (Trial/Dataset/Task setup, settings.phil, remote access, photon energy calibration). Covers cctbx.xfel.merge (full phil), cctbx.xfel.xtc_process, cctbx.xfel.process, cctbx.xfel.submit_job, cctbx.xfel.time_varying_refinement, cctbx.xfel.detector_residuals, cctbx.xfel.mpi_integrate, xfel.fee_calibration, and all other cctbx.xfel.* tools. Trigger on: cctbx.xfel, xfel.merge, xtc_process, time_varying_refinement, TDER, ensemble refinement, mpi_integrate, xfel submit_job, detector_residuals, cctbx SFX pipeline, merging.d_min, scaling.model, postrefinement, filter_experiments_by_rmsd, cctbx.xfel weather drift trumpet_plots, cctbx.xfel GUI, Trial

Skill.md ↗

writing-plans-codex

把已批准的设计文档转成实施计划并用 Codex 进行对抗式审查。

▸ 推荐理由将规划与自动化审查结合,适合追求快速交付的工程团队。
▸ 风险提示无人工门槛的自动执行链,存在失控或误执行风险。
开发编程计划生成对抗式审查CodexCodexClaude CodeSkill.md ↗

sdd-with-codex-implementer

在子代理驱动开发流程中由 Codex 作为实现者按任务写代码,Claude 负责控制与复核。

▸ 推荐理由把细粒度实现交给 Codex 执行,同时保留 Claude/人工审查以控制风险。
▸ 风险提示会自动化生成并应用代码变更,存在引入安全、质量或权限滥用的风险。
开发编程SDDCodex 实现者子代理任务执行CodexClaude CodeSkill.md ↗

pr-triage-codex

用 Codex 辅助将红色 PR 逐步整理并修复评审意见与 CI 故障直至通过。

▸ 推荐理由自动化汇总问题并批量生成修复建议,能显著减少人工 triage 成本。
▸ 风险提示会生成并应用代码修改,可能要求仓库写权限并存在引入错误或安全问题的风险。
开发编程PR 管理CI 绿化Codex 辅助问题分类CodexClaude CodeSkill.md ↗

codex-lanes

为 codex exec 调用提供统一的 lane、守卫与结果 schema 的参考配方。

▸ 推荐理由统一调用模式并内置失败守卫,减少 Codex-CLI 使用中的常见错误。
▸ 风险提示包含对本地 codex CLI 的调用与检查,可能需要执行本地二进制或高权限。
开发编程CodexCLI 规范守卫执行模版CodexSkill.md ↗

brainstorming-codex

在 Claude 产出后由 Codex 攻击/审核规格的双模型头脑风暴与设计流程。

▸ 推荐理由构建可审计的 Claude→Codex 设计—审查闭环,适合产品/工程启动场景。
▸ 风险提示会调用 Codex CLI 执行步骤,可能触发本地命令执行或外部调用风险。
开发编程头脑风暴Codex 审核双模型流程规范化CodexClaude CodeSkill.md ↗

youqing-laozhao-skill

接受漫画书名与作者,调研并产出中文长评网页与自动渲染的长 PNG。

▸ 推荐理由把研究、撰写到成品页面与长图一体化,适合需要成品交付的影评/书评场景。
▸ 风险提示调研可能引用受版权保护的资料或外部链接,存在版权与引用准确性风险。
内容创作漫画评测调研长图生成中文稿件通用Skill.md ↗

yaml-infographic

Create consistent single-canvas infographics from a structured YAML information architecture and a versioned global visual style. Use when the user asks for an infographic, information graphic, social-media long image, process graphic, comparison graphic, timeline, data story, checklist, anatomy callout, or a YAML-planned single image in 1:1, 4:5, 9:16, 16:9, or A4 proportions. Supports baked visuals and text-free plates with exact editable overlays.

Skill.md ↗