openJiuwen-ai
- 123 skills
- 0 followers
- 4 hours ago last updated
- ▌ Minimax PDF · openjiuwen-ai bundleUse this skill when visual quality and design identity matter for a PDF. CREATE (generate from scratch), FILL (complete form fields), or REFORMAT (apply design to an existing doc). Uses token-based design system.
- ▌ Skill Vetter · openjiuwen-ai bundleSecurity-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
- ▌ Canned Responses · openjiuwen-ai bundle为常见法律询问生成模板回复，并识别何时需要个别关注。适用于回复数据主体请求、供应商询问、NDA请求、发现保全等日常法律询问，或管理回复模板时使用。
- ▌
- ▌ Factcheck · openjiuwen-ai bundle对给定内容进行事实核查，拆解主张、多源验证、真伪裁定。适用于验证文章/报告/演讲中的事实性陈述、数据准确性、因果关系的可靠性。触发词包括'事实核查'、'验证真伪'、'核实数据'、'这个说法对吗'、'查证'。不适用于纯主观观点讨论、文学创作评价、或仅需1-2次搜索即可确认的简单事实。
- ▌
- ▌
- ▌ Official Doc Formatter · openjiuwen-ai bundleFormat Word documents to official Chinese government and enterprise document standards, including requests, plans, attachments, heading hierarchy, spacing, fonts, and punctuation rules. Use when users need to normalize or restyle formal DOCX materials, especially official memos, requests, plans, notices, or attachment documents that must match GB/T-style formatting conventions.
- ▌ Lidan Writing Framework · openjiuwen-ai bundleApply Li Dan's seven-step explanatory writing framework to turn complex ideas into clear, teachable long-form writing with stories, contrasts, examples, and reading lists. Use when users want structured knowledge writing, concept explainers, educational essays, public-intellectual style articles, or a repeatable framework for drafting deep but accessible content.
- ▌ Knowledge Organizer · openjiuwen-ai bundleAn OpenClaw- and Codex-compatible knowledge organization skill for importing articles, organizing notes, syncing to Feishu and Tencent IMA, applying tags, archiving content, generating summaries, and suggesting related notes.
- ▌ CI · openjiuwen-ai bundleRead, diagnose, and change the CI pipelines: CodeArts on GitCode merge requests and GitHub Actions on the mirror. Use when a pipeline or job fails, when editing .codearts/workflow/ or .github/workflows/, when asking which platform runs which layer, when reproducing a pipeline failure locally, or when a job needs the bubblewrap sandbox. Running the layers before a merge request and reading the result comment on it belong to the create-pr skill.
- ▌ Code Engineer · openjiuwen-ai bundleUse this skill when you need to write and execute Python/R code to process, transform, and analyze data, delivering reproducible computational results with complete code-level methodology documentation. Supports statistical analysis, data transformation, visualization, method justification, and structured result output.
- ▌ Evolve Design · openjiuwen-aiUse when the user wants to improve something by repeated search rather than one edit — a program, a prompt, a document, a pipeline, a configuration, an experimental protocol. Triggers on `/evolve-design`, "把这个做得更好", "搜索一个更好的方案", or any request to optimise against a measurable target. Runs a short checkpointed design conversation, verifies the scoring can rank candidates, then calls `create_evolve_run`. Not for a single fix, a refactor, or a question about existing code.
- ▌ Report Writer · openjiuwen-aiReceives Domain Summaries (Knowledge Summary, Analysis Summary, or both) from the science-research-team orchestrator, performs cross-domain synthesis (identifying SUPPORTS/CONTRADICTS/INFORMS/INDEPENDENT connections), surfaces contradictions verbatim, and produces a formatted output per user requirements (or the default 7-section template if no format specified). NOT for performing research, conducting analysis, evaluating results, or resolving contradictions.
- ▌ Deploy · openjiuwen-aiDeploy or run the ScienceDiscovery stack, either as host processes (scripts/start-stack.sh) or with Docker Compose. Use when the user asks to deploy, start, run, or containerize ScienceDiscovery, or asks for the UI URL, SSH port forwarding, or start/stop/restart/log commands.
- ▌ Idea Tree Team · openjiuwen-ai bundlePrepare supplied inputs for autonomous Idea Tree research and explain its saved results.
- ▌ Gitcode · openjiuwen-ai bundleOperate GitCode issues, PRs, wikis, code/MR refs, and cached org templates. Use for gitcode.com, atomgit.com, GitCode issues/PRs, wiki, or /gitcode.
- ▌ Result Evaluator · openjiuwen-aiEvaluate analysis results for quality and reliability. Scores Accuracy, Completeness, Robustness, and Relevance (0-10), checks source reliability and methodology, audits statistical rigor, and decides ACCEPT_AND_PROCEED or REVISE_AND_RETRY. NOT for performing analysis or modifying results.
- ▌ Create Pr · openjiuwen-aiOpen a merge request on GitCode: run the UT/ST/E2E layers locally first, write a body that says what was verified with numbers, target openJiuwen/sciencediscovery, then read the bot comments and the CodeArts result the merge request receives. Use when asked to create a PR or MR, submit a change for review, when a branch is ready to propose, or when interpreting a merge request's CI result comment or ci-* labels.
- ▌ Citation Reviewer · openjiuwen-aiVerify cited paper identity and decide whether claims in an Artifact are supported by their linked Evidence using a governed source snapshot.
- ▌ Evidence Extractor · openjiuwen-aiUse this skill when a research workflow needs source-grounded evidence extraction from a `literature_sources.json` package or compatible source list. Produces an integration-ready evidence package with claims, evidence, citations, confidence/strength labels, quality notes, and partial results when some sources cannot be processed. Not for literature search, deduplication, final report writing, cross-source synthesis, or workflow coordination.
- ▌ Insight Aggregator · openjiuwen-ai bundleSynthesize insights from multiple independent assessment results into a concise, actionable summary. Cross-validates expert scores, identifies patterns and discrepancies, and produces a semantic insight that propagates upward through the Idea Tree. Does NOT calculate or override scores.
- ▌ E2e Testing · openjiuwen-aiDesign, extend, and run ScienceDiscovery E2E user journeys through the browser, public HTTP API, CLI, or local product stack. Use when changing user-observable behavior, selecting mocked/real journeys, isolating a test stack, reporting user outcomes, or attributing an E2E failure. Browser journeys use the pinned Playwright environment in test/.
- ▌ Literature Searcher · openjiuwen-ai bundleUse this skill when a research workflow needs verified academic source retrieval through literature-search MCP interfaces available in the current session before evidence extraction. Discovers and selects relevant interfaces adaptively, then produces deduplicated literature JSON and coverage notes. Not for full-text reading, evidence extraction, final report writing, or workflow coordination.
- ▌ Assessment Screening · openjiuwen-ai bundleIndependent multi-dimensional assessor for material design candidates. Loads configurable scoring rubrics to evaluate candidates from different expert perspectives. Each dispatch produces structured scores, pros/cons, and verification data without sharing results with other assessors.
- ▌ Computation Reviewer · openjiuwen-aiCheck whether numeric claims marked with Evidence aliases in an Artifact agree with the linked Evidence content and context.
- ▌ Science Research Team · openjiuwen-ai bundleUse this skill to orchestrate a multi-domain research team for literature/evidence research and data analysis. Dispatches the right domain(s) based on the user's request, runs each domain to completion with its own loop, then synthesizes into reports. Load when the user needs rigorous research with knowledge or data analysis.
- ▌ Creative Material Design · openjiuwen-aiCreative material design specialist for water treatment materials. Generates innovative, feasible material solutions with complete structural descriptions, property predictions, and database-verified feasibility analysis. Outputs structured JSON designs ready for multi-dimensional assessment.
- ▌ Antibody Protenix Pipeline · openjiuwen-ai bundlePrepare, submit, monitor, and summarize the real RFdiffusion to ProteinMPNN to Protenix antibody pipeline through the ScienceDiscovery Runner NPU Broker, including custom workspace-local antigen PDBs and chain-labelled epitopes.
- ▌ Life Science Evidence Brief · openjiuwen-ai bundleResearch scientific questions with enabled literature and database connectors, distinguish curated annotations from paper evidence, and produce a cautious claim-to-source brief. Use for gene/protein function summaries, literature scans, evidence tables, paper abstract reading, or cited scientific reports.
- ▌ Structure Pocket Inspection · openjiuwen-ai bundleInspect a local PDB structure, summarize chains and residue composition, and identify protein atoms near a user-specified ligand or pocket center. Use for reproducible first-pass structure and binding-site inspection before heavier modeling.
- ▌
- ▌ Python Patterns · openjiuwen-aiPython idioms, immutability, async patterns, and anti-patterns for agent-core.
- ▌
- ▌ Verification Loop · openjiuwen-aiFormalizes agent-core's make check/type-check/test/fix pipeline into a structured 6-phase verification skill.
- ▌ Github Com · openjiuwen-ai bundleNavigate github.com in mobile Chrome to retrieve repository information (files, README, contributors, languages, releases, packages) and switch to desktop site mode for richer metadata (forks, stars). Use for in-browser research, not git CLI or GitHub API operations.
- ▌ Scheduling · openjiuwen-ai bundleComplete calendar-and-clock workflows on mobile (find events, read start times, create alarms or reminders relative to meetings). Use when the user asks to schedule, set an alarm before/after a meeting, find appointment times, or coordinate Calendar with Clock.
- ▌ Booking Dot Com · openjiuwen-ai bundleComprehensive step-by-step workflow for navigating the Booking.com mobile site to search for destinations, select dates/occupancy, browse results, and extract detailed hotel information including pricing and facilities.
- ▌
- ▌
- ▌ Skill LLM Judge · openjiuwen-aiAudit Agent Skill design quality through static analysis of SKILL.md — no prompt execution, no code running. Scores 7 design dimensions (100pts): knowledge ratio, expert knowledge craft, specification compliance, progressive disclosure, pattern + freedom fit, predicted usability, and output specification. Use when reviewing a skill's DESIGN before or after functional testing. Outputs structured design_score.json for eval pipeline consumption. Do NOT use when you want to run the skill on real prompts or measure actual output quality — use skill_tester for that.
- ▌ Skill Tester · openjiuwen-ai bundleGenerate and run synthetic test cases against any skill by executing real prompts through the skill and grading actual outputs — functional and behavioral testing only. Use when verifying a skill WORKS: correct outputs, edge case handling, error recovery, end-to-end workflow. Triggers on: "test this skill", "generate test cases", "run evals", "validate the skill", "run tests on skill", or when given a skill path to verify. Outputs pass_rate and structured assertions.json for eval pipeline consumption. Do NOT use for reviewing skill design quality, knowledge delta, anti-patterns, or SKILL.md structure — use skill_llm_judge for that.
- ▌
- ▌ Verify Ext · openjiuwen-aiRuntime extension 验证规范 — 验证 harness package 中 tools、rails、skills 是否能真实热加载并可运行
- ▌
- ▌
- ▌ Generate Skill From Url · openjiuwen-ai bundleGenerates a multimodal Skill markdown file from a source URL. Use this Skill when a user wants to turn a web tutorial, product support article, or software guide into a reusable agent Skill with concise steps and embedded screenshots.
- ▌ Skill Safety Eval · openjiuwen-aiEvaluate Agent Skill safety through static analysis of SKILL.md instructions — no prompt execution, no code running. Gates the eval pipeline at score_pct ≥ 0.80: skills that fail do not ship regardless of design or functional scores.Scores 5 safety dimensions (100pts): harmful instruction potential, scope containment, data handling safety, injection resilience, and guardrail presence. Use when auditing whether a skill's instructions could cause an agent to produce harmful outputs, exceed appropriate scope, mishandle sensitive data, or be hijacked via injection. Do NOT use for design quality review (use kill-llm-judge) or functional testing (use skill-bench).
- ▌ Eval Pipeline · openjiuwen-aiFull skill evaluation pipeline — the single entry point for comprehensive skill assessment. Orchestrates multiple evaluators in sequence, aggregates their scores, and produces a final weighted verdict. Use when you want to fully evaluate a skill end-to-end: design quality, functional correctness, and (when available) safety. Triggers on: "evaluate this skill", "run full eval on skill", "assess skill quality", "is this skill ready to ship". Do NOT use when you only want design review (use skill_llm_judge directly) or only want to run functional tests (use skill_tester directly).
- ▌ TS Plan · openjiuwen-aiDecide the paper's title from the run's evidence. Use once, at the very start of a reporting session, before ts-write.
- ▌ TS Latex · openjiuwen-ai bundleAssemble and compile the final PDF, repairing compile errors. Use last, after ts-review.
- ▌ TS Write · openjiuwen-aiDraft every section's LaTeX body from the evidence. Use after ts-plan, before ts-review.
- ▌ TS Figure · openjiuwen-ai bundlePlan, generate, integrate, and verify the figures needed to communicate the paper's scientific claims. Use after ts-write, before ts-review.
- ▌ TS Review · openjiuwen-ai bundleCheck every section against deterministic rules, then run a three-angle adversarial review and fix what survives. Use after ts-write, before ts-latex.
- ▌ UI E2e · openjiuwen-ai bundle运行 JiuwenSwarm Web UI 端到端测试并收集截图、日志、report.md、report.json。用于验证任务多模态、Todo 和 Cron Web UI 流程、复现浏览器交互问题、选择运行解释器、准备 Playwright 环境，或返回可操作的失败证据时。
- ▌ XLSX · openjiuwen-ai bundleOpen, create, read, analyze, edit, repair, or validate spreadsheet files (.xlsx, .xlsm, .xltx, .csv, .tsv) with zero format loss, with first-class support for Chinese / CJK content. Use whenever the user asks to create, build, modify, fill, analyze, read, repair, validate, or format any spreadsheet, workbook, report, dashboard, budget, financial model, pivot table, or tabular data file. Handles Chinese text, encodings (UTF-8 / GB18030 / GBK), full-width characters, CJK fonts in rendering, and Chinese number/date conventions (¥, 万/亿, 年/月/日). Covers: creating new workbooks from scratch, reading and analyzing existing files, editing existing workbooks while preserving pivot tables / VBA macros / charts / sparklines / conditional formatting / named ranges, formula recalculation and validation, visual review of results, and applying professional formatting. Triggers on 'spreadsheet', 'Excel', 'workbook', '表格', '工作簿', '.xlsx', '.xlsm', '.csv', 'pivot table', '透视表', 'macro', '宏', 'chart', '图表', 'financial model', '财
- ▌ DOCX Pro · openjiuwen-ai bundleProfessional Word (.docx) document generation, faithful text replacement, Markdown round-trip conversion, document inspection, TOC insertion and watermarking. Use when the user asks to create rich-formatted Word documents (cover page, TOC, multi-level headings, styled tables with merged cells, images, code blocks, quotes, headers/footers, page numbers, watermarks), replace or batch-replace text in existing docx files while preserving every other byte and all formatting (handles text split across runs, tables, headers/footers), convert Markdown to docx or docx to Markdown, inspect document structure, insert TOC fields, or add text watermarks to existing docx files. Triggers on docx-pro, 生成 Word 文档, 生成报告文档, 富格式文档, Word 替换, Word 文本替换, Word 保真替换, 批量替换, 合同替换, 替换甲方, Markdown 转 Word, Word 转 Markdown, md 转 docx, docx 转 md, 插入目录, 添加水印, 文档结构分析, 专业排版.
- ▌ LLM Wiki · openjiuwen-aiBuild and maintain a persistent knowledge base (LLM Wiki) using native backend tools. Supports ingesting PDFs, Markdown, and TXT files, and querying or linting the data. Use this whenever the user wants to add, retrieve, or manage their own local knowledge base and documents.
- ▌ Skvm Jit · openjiuwen-ai bundleTrigger `skvm jit-optimize` with post-task evidence so it generates a reviewable optimization proposal for a skill. Use after finishing a task that was driven by any skill (not just SkVM-compiled ones) when the outcome was a failure, was partial, or the skill's instructions clearly caused confusion or an unnecessary detour. Clean successful runs usually do not need an optimization pass. For general skvm usage (profile/aot-compile/bench/proposals), use the sibling `skvm-general` skill instead.
- ▌ Akg Agents · openjiuwen-ai代理执行 AKG 算子任务。先检查固定仓库与分支；若 `~/.akg/check_env.md` 不存在则强制 `FULL_SETUP=true`；忽略所有 `akg_cli` 检查和使用；后端代码生成直接执行 `run_workflow.py --workflow kernelgen`。
- ▌
- ▌ Gitcode API · openjiuwen-ai bundle使用 gitcode-api 命令行工具访问 GitCode REST API。适用于需要查询仓库、Issue、Pull Request、文件内容、用户、组织、搜索结果，或需要获取 Issue 与 Pull Request 模版的场景。还可用于总结仓库最近 commit / PR / issue、责任人、模块、风险或对外接口影响。
- ▌ Ppt Creation · openjiuwen-ai bundle当任务要求创建、读取、解析、提取、编辑、修改、合并、拆分、转换或检查 PPT / 幻灯片 / 演示文稿 / deck / slides / presentation / .pptx 文件时使用。本 skill 以内置模板为版面底座，让 Agent 在内容区自由设计；支持原生 PptxGenJS 绘制、可选组件、用户素材、论文原图和合规网络素材。
- ▌ Skvm General · openjiuwen-aiDrive the skvm CLI on behalf of a user to profile models, AOT-compile skills, run skill-assisted tasks, run benchmarks, and manage compiled proposals. Trigger when the user asks to "profile", "aot-compile", "bench", "run a single ad-hoc task with a skill", or asks about skvm proposals. Do NOT trigger for `jit-optimize` or when the user wants to optimize/improve a skill — use the sibling `skvm-jit` skill instead.
- ▌ Agent Creator · openjiuwen-ai bundleAgent 模板包的创建与更新器：从自然语言创建可被 JiuwenSwarm 加载的 agent 模板包，或在同一流程的 update mode 下修改已有自定义专家。 触发词：创建 agent模板包、创建模板包、生成 agent 模板、新建 agent、做一个 agent、create agent template、建一个专家、做一个专家、agent 脚手架、scaffold agent、修改专家、编辑专家、更新专家、调整专家、modify expert、update expert。当用户说「帮我做一个能 XX 的 agent」「创建一个 XX 专家」「改一下 XX 专家」「更新已有专家」时，务必使用本 skill。
- ▌ Local Doc Ocr · openjiuwen-ai bundle本地离线 OCR——将扫描件 PDF/图片提取为文字。双引擎：(A) RapidOCR 纯本地离线一键出 .md（默认保底、不联网、不占 token、可批量）；(B) 支持多模态读图能力的 Agent 逐页读图（高质量，擅长表格/手写/印章，如 WorkBuddy 的 Read 工具）。全程本机处理，适合涉密/敏感与隐私资料。触发词：OCR、识别、读PDF、读一下这份、把这个读出来、转成文字、扫描件提取、图片转文字、识别扫描件、PDF转文字、手写识别、提取图片里的字、这份文件读不出来、资料入库（如喂给知识库）时遇到扫描件。注：用户说"识别隐患/安全隐患"属安全检查类技能，本技能只做文字提取。
- ▌ Skill Creator · openjiuwen-aiUnified entry point for all Skill creators. Routes skill creation and editing requests to the correct specialized creator. Use whenever the user wants to create a new Skill, edit an existing Skill from chat, or when the conversation scene is create_skill / edit_skill. Chooses among skill-creator-normal, swarmskill-creator, and skill-omni-creation, then ensures the final Skill package is delivered via send_file_to_user (never silently installed to workspace). Do not bypass this entry by calling specialized creators directly for create_skill / edit_skill scenes.
- ▌ Plugin Creator · openjiuwen-ai bundlePlugin 插件包的创建与更新器：从自然语言创建可被 JiuwenSwarm 加载的 plugin 插件包，或在同一流程的 update mode 下修改已有自定义插件。 触发词：创建插件、创建 plugin、做一个插件、plugin 脚手架、修改插件、编辑插件、更新插件、调整插件、create plugin package、插件包、帮我做一个 XX 插件、给 agent 加 XX 能力包。当用户说「帮我做一个 XX 插件」「创建一个能力扩展包」「改一下 XX 插件」时，务必使用本 skill。
- ▌ Dynamic Memory CLI · openjiuwen-ai bundleOperate a dynamic, UT-backed agent memory with atomic Pending publication, merged Pending/Built search, Memory Snapshot and history cursor state, Built-only validation, and exact-content migration. Use when an agent Harness must extract memory asynchronously while foreground work continues, publish immediately searchable Pending UTs, build them into a stable index, or inspect and recover an eternal-conversation memory project.
- ▌ Project Maintainer · openjiuwen-ai bundleUse when initializing or updating structured in-repository project-maintenance docs; exploring existing code; analyzing a full project or repository; documenting complete source symbol coverage, coverage maps, symbol audit maps, or health for every top-level class, top-level function, and class method; summarizing git history; using project-maintainer as maintenance-aware context during a bug fix, feature change, or refactor; or syncing docs after verified code changes.
- ▌ Swarmskill Creator · openjiuwen-ai bundleCreates, converts, or modifies Swarm Skills — the multi-role (多角色团队) extension of the Skills standard, optionally with an executable SwarmFlow orchestration script. Use when building or refactoring a multi-agent team, generating workflow (工作流/编排) orchestration code, or upgrading a single-agent skill into a collaborating team. Do NOT use for ordinary single-agent skills — use create-skill instead.
- ▌ Delayed Restart App · openjiuwen-ai bundle安排延迟重启本 Agent 所在的服务（JiuwenSwarm app）。执行后当前 Agent 进程会被终止并重新启动，当前会话会断开。用于用户要求重启、配置更新需生效、或服务异常需重载时。使用 bash 调用脚本。
- ▌ Harmonyos Dev Suite · openjiuwen-ai bundleUnified HarmonyOS development entry skill for ArkTS, ArkUI, DevEco CLI, build/run/debug, device logs, testing, multi-device adaptation, atomic services, native development, stability diagnosis, and optional HarmonyOS atomic Skill routing. Use when the user mentions HarmonyOS, OpenHarmony, HongMeng, 鸿蒙, ArkTS, ArkUI, DevEco, .ets, oh-package.json5, build-profile.json5, HAP/HAR/App, emulator, hilog, hdc, hvigor, or HarmonyOS app development.
- ▌ Skill Omni Creation · openjiuwen-ai bundle当用户说「从这个链接/URL生成skill」「把这个网页/教程做成skill」「从这个视频提取步骤」时触发。先读取此 SKILL.md，再按步骤调用 scripts/ 下的 Python 脚本完成：爬取网页或下载视频 → 下载图片/抽帧 → 生成标准 Skill Markdown 文件；若内容需要编程实现，自动编写并验证配套脚本（scripts/）。
- ▌ Skill Creator Normal · openjiuwen-ai bundleCreate new single-agent skills, modify and improve existing single-agent skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
- ▌
- ▌ Openjiuwen Deepsearch · openjiuwen-ai bundle知识增强型深度检索与深度研究，支持查询规划、信息收集、理解反思、研究报告生成等多 Agent 协同。使用场景：金融分析研报、学术与政策研究、企业级深度搜索等复杂推理任务，可以生成Markdown、Doc和Html格式的研究报告。**每次使用该技能之前都先完整阅读一遍SKILL.md学习技能。**
- ▌ Deepep To Cam Converter · openjiuwen-ai bundle识别代码中可替换为CAM算子的DeepEP算子，基于实际运行参数校验约束，自动完成通信域转换（NCCL->HCCL）、设备适配（CUDA->NPU）及算子替换。
- ▌ Huawei Cloud Maas Setup · openjiuwen-ai bundle引导用户完成华为云 MaaS 服务购买和 API 配置。脚本自动完成授权更新、API Key 创建、模型开通；用户只需完成登录、充值、选择要开通的模型。整个流程预计 3-5 分钟。 适用于首次运行配置、华为云 MaaS 服务开通、API 凭证获取、模型服务配置等场景。 当用户提到华为云、MaaS、Maas、开通模型服务、配置 API、首次配置、引导配置、 一键配置等意图时触发。
- ▌ Financial Document Parser · openjiuwen-ai bundleExtract and analyze data from invoices, receipts, bank statements, and financial documents. Categorize expenses, track recurring charges, and generate expense reports. Use when user provides financial PDFs or images.
- ▌ Skill Gen 4 Enterprise Doc · openjiuwen-ai bundleTurn SOPs, URLs, or stated intent into an installed skill package. Use when the user asks to generate a skill from a document, link, or workflow. Follow SKILL.md, reference/sop-structure-pipeline.md, and reference/generator-worker-spec.md; finish by promoting the draft into the runtime skills directory (see reference/operator-playbook.md) in the same turn.
- ▌
- ▌ Aet Req Design · openjiuwen-ai bundleRequirements design skill - transforms requirements analysis specifications into implementable system design specifications through architecture design, module change planning, interface design, and DFx strategy. Use when: (1) you have completed requirements analysis and need to produce a requirements design specification, (2) you need to clarify how requirements integrate into the existing system, (3) you need module-level change planning with frozen zones and interface contracts, (4) you need to produce a SDD-style design document, or any requirements design and architecture design tasks.
- ▌ Aet Req Review · openjiuwen-aiGate pipeline for requirement deliverables — enforces automated quality gates and interactive user validation before handover. Trigger when: (1) structured PRD/spec/SDD reviews are required, (2) executing a Stage demanding review-then-revision cycles, (3) workflows necessitate HCritic automated checks coupled with user-in-the-loop revision.
- ▌ Ascend Moe Optimizer Auto Trace · openjiuwen-ai bundle为昇腾算子在源码中接入 TRACE_POINT 与 MoeTracing，串通 trace_preprocessor、profiling tensor、point_map.json、 save_profiling_data 与 trace_collector 生成 Chrome trace。强调门禁 G1–G5：全链路预处理与 OPP、profiling 为数据输出最后一位、 整条编译与示例脚本联调、落盘路径在 spawn 前 resolve。遵循函数级粒度与就地扩展，禁止另注册 xxx_profiling 类第二入口， 保持原 Op 与 torch.ops 名称及签名不变。在用户提到算子打点、Profiling、Chrome trace、MoeTracing，或将结论写入本 skill 时读取。
- ▌ Cross Channel History Retrieval · openjiuwen-ai bundle跨会话检索聊天原文（记忆不足时再用）。在回答任何关于历史事件、日期、人物、过去对话的问题时，如果记忆中没有相关信息或不足以回答，则需要使用跨会话检索聊天原文。用 mcp_exec_command 执行 scripts/search_history.py，读 ~/.jiuwenswarm/agent/sessions/*/history.json。支持 channel、session_id、关键词、时间窗。如果搜索结果不足，尝试用不同的关键词再次搜索。
- ▌ Aet Req Analysis · openjiuwen-ai bundleRequirements analysis skill - transforms raw requirements into structured specifications through Socratic dialogue, behavior analysis, and requirement specification design. Use when: (1) requirements are unclear or need decomposition, (2) you need to produce a requirements analysis specification from user input, (3) you need structured functional and non-functional requirements with priority labels, (4) you need acceptance criteria and test case definitions, or any requirements clarification and specification generation tasks.
- ▌ Ascend Moe Optimizer Trace Analyzer · openjiuwen-ai bundle在用户提供 Chrome/Perfetto trace.json、或排查 Ascend 上 MoE/FusedDeepMoe 等算子性能时使用。按 phase、category、core group、tid 统计耗时、overlap、bubble，输出 CSV、Markdown 报告与确定性诊断；可选外部 LLM 扩写分析。默认 phase 映射面向 UMDK FusedDeepMoe，其它 trace 需替换或扩展 config/phase_map.yaml。
- ▌ Skill Name 1 · openjiuwen-ai对 Leader 分派的专业问题进行结构化分析，返回有证据支撑的结论、不确定性和下一步建议。用于专家团 Member 收到需要独立研究、评估或方案建议的子任务时。
- ▌
- ▌ Devflow Tdd · openjiuwen-ai bundle在实现任何功能或修复任何缺陷、即将编写实现代码时使用；设计确认后的整个实现期都适用。强制测试先行的 RED→GREEN→REFACTOR 循环。不用于规格编写、设计决策或纯文档修改。
- ▌ Devflow Ship · openjiuwen-ai bundle在工作项全部任务完成、测试与代码评审闭环后收尾时使用：对照 Definition of Done 核验完成证据、把已确认的规格与设计沉淀为长期文档资产（promotion）、写 closeout 记录。不用于实现、评审或修复。
- ▌ Using Devflow · openjiuwen-ai bundleDevFlow 工作流的入口。在以下情况使用：开始一个新的开发任务、不确定当前该做规格/设计/实现中的哪一步、需要从已有工件恢复进度、或用户提到 DevFlow / 规范驱动开发 / 高质量开发流程时。
- ▌ Devflow Design · openjiuwen-ai bundle在规格确认后、写代码前做软件设计时使用；也在设计评审被打回、或实现中发现模块边界/接口契约/错误处理需要重新设计时使用。涵盖模块划分、接口契约、错误模型、数据所有权、方案取舍与测试设计。
- ▌ Devflow Review · openjiuwen-ai bundle在规格、设计、测试或代码需要独立评审时使用：阶段产物完成后的把关、人要求 review、或对既有产物做专项检查时。评审必须由作者之外的独立上下文执行，产出 findings 与 verdict。
- ▌
- ▌
- ▌ Caveman Install · openjiuwen-aiInstall and configure the caveman plugin for Claude Code on the current machine. Trigger when user says "install caveman", "setup caveman", "配置 caveman", "安装 caveman", or when caveman hooks/statusline are missing from settings.json. Also applies when user wants to verify, troubleshoot, or uninstall caveman. Not applicable to: using caveman mode (that's the built-in caveman skill), general Claude Code configuration unrelated to caveman.
- ▌ Devflow Specify · openjiuwen-ai bundle在开始一个新的开发工作项（功能、变更、需要正式规格的缺陷）时使用，把一句话需求或散乱输入澄清成可测试的规格；也在规格评审被打回需要修订时使用。不用于产品方向决策或实现设计。
- ▌