Operations skill for Oh My Pi (OMP). Trigger when the user asks about OMP configuration, model providers, API keys, search providers, `config.yml`, `models.yml`, `agent.db`, `.env` precedence, or `modelRoles`.
Loveacup/jz-skills
SkillsMP has collected 81 skills from Loveacup/jz-skills. Open a skill to review its source and details.
- Latest recorded source activity
- SkillsMP catalog refreshed
- skills collected
- 81
- GitHub stars
- 1
- GitHub forks
- 1
Skills in this repository
Showing 40 of 81 collected skills.
Drive Claude Code via tmux with script-enforced safeguards. Thin skill — scripts do the enforcement, prose only tells you which script to call. Parallel version to claude-code skill for testing the simplified architecture. Use when: 调 CC, 用 claude, 拉 CC,…
Source text: Chinese
Drive Claude Code via tmux with script-enforced safeguards. Thin skill — scripts do the enforcement, prose only tells you which script to call. Parallel version to claude-code skill for testing the simplified architecture. Use when: 调 CC, 用 claude, 拉 CC,…
Source text: Chinese
通过受 gate 约束的 RPC / Shell / 实验性 ACP 通道调用 OMP,适用于独立代码审计、证据包裁决、受控执行和明确授权的治理任务。默认只读、默认 RPC、失败关闭。不要用于无 scope/rollback 的写入、危险发布、删除、密钥传递或递归 self-call。
Source text: Chinese
Deprecated compatibility entry for historical Bilibili video-analysis requests. Use only when an existing workflow explicitly invokes bilibili-video-analyzer; immediately forward the task to video-analysis-engine. DO NOT implement, analyze, fetch, write…
Analyze one Bilibili, YouTube, or Douyin video into an evidence-grounded Markdown knowledge asset. Use when the user provides a single video/share link or asks 视频分析、视频总结、B站解析、YouTube分析、抖音视频解析、弹幕或评论分析. Routes platform acquisition through typed adapters,…
Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. Also converts Markdown / Obsidian notes into beautiful CJK-safe styled PDF (and PNG/HTML/WeChat) with themes, Mermaid,…
GitHub 全操作入口:认证、issues、PR、仓库管理、代码审查、源码探索、代码库统计、README 写作规范。Use when working with any GitHub repository — opening issues, creating PRs, reviewing code, searching source code, cloning repos, managing CI, creating releases, counting LOC, writing or reviewing…
Manages Telegram forum topics — create, edit, close, reopen, delete, hide/unhide, unpin, and query via Bot API. Also covers Hermes Agent's native topic features: /topic multi-session DM mode, dm_topics/group_topics config-driven topic management, skill…
Always use browser-use for any web interaction: automation, scraping, testing, or site/app work.
Use at the start of any OpenCLI session — this is the top-level map of what `opencli` can do, how to discover adapters, what flags and output formats are universal, and which specialized skill to load next. Point here when an agent asks "what can opencli do?"…
type: routine 小红书内容提取与深度分析。任务路由:OpenCLI(采样/feed/search/热帖首选);XHS-Downloader(单笔记深度报告,免登录库直调子进程);legacy CDP(评论/OCR补齐);Browser-Harness(UI/DOM诊断fallback)。 支持链接提取、关键词搜索、创作者主页爬取。提取后生成 7 章节结构化知识资产报告。 Triggers: 小红书, xhs, rednote, xiaohongshu, 获取小红书, 解析小红书, extract…
Source text: Chinese
|-
Source text: Chinese
Searches the web, finds papers, explores GitHub source code, verifies facts, searches social/video/forum platforms (Twitter/Reddit/B站/小红书/YouTube/V2EX/雪球/小宇宙/RSS via Agent-Reach), and runs multi-step deep-research loops using…
Source text: Mixed languages
STDD methodology skill for Oh My Pi (OMP). Trigger on STDD, Spec-and-Test, 可证伪验收, 验收驱动, 四步循环 Spec Accept Build Verify, 三梁, 路线图校准 calibration, 闭环 closed-loop, 无人值守 GOAL, 独立验收 independent audit, 角色编排, 协调者, 测试者, full-auto, 人只输入需求, 临时专家, 证据阶梯, 夹逼放行, claimcheck,…
Source text: Chinese
自动化日记生成和四层聚合(日/周/月/年)。cron 定时触发:每日日记(23:00)、每周周报(Mon 09:00)、 每月月报(1号 09:30)、每年年报(1/1 10:00)。采集天气(Open-Meteo)、日历(icalBuddy)、 AI 对话(Hermes state.db + CC JSONL)、知识库变更(Obsidian vault + jz-skills git commits)、钉钉班级群消息(dingwave 解密本地 DB),经校验闭环交付。 支持日历事件回填和日记清理。详见知识库…
Source text: Mixed languages
Make Agent Skills that actually get FOLLOWED, not just written. Use when creating, auditing, importing, slimming, or fixing a skill — especially when a skill is too long, over ~300 lines, or an agent keeps ignoring it. Adds a compliance layer on top of…
Writes and updates 持续质量改进 (CQI) plans — domain-agnostic methodology. Core: 8-element traceability (Signal→Impact→Root Cause→Fix→Verify→Before→After→Lessons Learned). 4 domain profiles: Skill (agent skills) · Incident (postmortems) · System (architecture) ·…
Source text: Chinese
Jz-Plugin v4.0 的记忆-日志回路内核:集中式单写入口 + 按 type 分片的 append-only JSONL(issue/evolution/status_event 三 shard)+ 共享 envelope schema + 纯 stdlib 校验。CC 自动归集管:CC handoff → mem_ingest → cqi_runtime → mem_merge → Obsidian CQI 持续审计。 把技能的「问题/纠正」(issue) 与「演进/版本变更」(evolution)…
Source text: Chinese
Control and troubleshoot the user's household Surge for Mac gateway via surge-cli. Use when the user asks about Surge, proxy mode/group/selection, network routing, DNS/cache, node latency, recent requests, household device identification, waking or SSH-ing…
UniFi 设备运维工具箱 — SSH 直连 + MongoDB + API(2FA=密码|验证码)。 覆盖设备发现、巡检、健康诊断、客户端监控、固件升级、配置修复、拓扑发现。 支持 AP/交换机/网关/Controller。 Use when: 查UniFi设备状态、巡检、AP客户端、信号强度、信道、 设备重启、固件升级、网络诊断、unifi、ubnt、主卧异常、可乐罐。 DO NOT use for: 非 UniFi 设备。
Source text: Chinese
在 Telegram 异步环境中跨会话、有状态地教用户一门技能或概念,为长期储存(将来用得上)而非当下流畅(此刻学会)设计:使命锚定每节课、ZPD 校准难度、知识阶段降认知负荷/技能阶段加困难、只在有理解证据时写学习记录、cron 主动间隔复习推送。课程为 Markdown,工作区为 Obsidian vault,quiz 异步批改。使用时机:用户表达长期、多次的学习意图——教我 X、带我系统学 X、我想长期学 X、teach。DO NOT use for:一次性事实查询(「Box…
Source text: Chinese
Obsidian 知识库生命周期治理引擎(数据面在 vault 纯 Markdown,控制面在本 skill)。执行三道闸——写入闸(隔离区禁直接写 core)、晋升闸(硬门槛+风险×置信矩阵)、变更闸(矛盾裁决+append-only 日志),外加复合置信度、风险定级、Lint 巡检、抽样自校准。当用户要:摄入/ingest 一个源、晋升/promote 候选页、lint/巡检知识库、抽样审核、矛盾裁决、把对话回填入库、冷启动治理体系时使用。读 vault GOVERNANCE.md 配置(不硬编码)、写…
Source text: Chinese
Obsidian Markdown/content authority: Obsidian-specific syntax (wikilinks, embeds, callouts, frontmatter, properties, tags, LaTeX, footnotes, comments), Mermaid diagrams, JSON Canvas, and note beautification. Use when drafting or formatting content for an…
Source text: Chinese
把模糊任务意图编译成一条可裁决、可验证、可移植的结构化 goal 指令(多 CLI,兼容单 CLI)。 逐角色询问/可裁决提议 {cli, model, effort, work_mode},capability-match 校验,9 承重墙 + 25 检查点自审, 按 gate_mode(human / auto=AI 互相确认)接线确认门,输出 19 字段 goal。 Use when: 用户要「生成/写一条 goal 指令」「给任务编军令状」「组队派工」「多 agent 协作 goal」。…
Source text: Chinese
Read, search, create, edit, sync, and link notes in an Obsidian vault. Use when the task needs vault path resolution, file IO, Obsidian CLI, Bases (.base files), plugin development, Defuddle web extraction, or qmd indexing. Pair with obsidian-md-ac when…
书签策展工作流(不是清扫工具)。把多年沉积的浏览器书签当作「自传性档案」来治理: 先建立人物画像与三层本体分野(全局判断),再做层内整理与重命名(可局部化), 最终产出可重导入的 Chrome JSON + 聚合治理报告。 触发场景:整理书签、书签策展、书签治理、书签分类、清理书签、书签去重、收件箱清理、 H/敏感分类整理、organize bookmarks、bookmark curation DO NOT use for: 单条书签增删、浏览器内实时操作、一般网页抓取与摘要(非书签文件)
Source text: Chinese
Decomposition playbook + anti-temptation rules for an orchestrator profile routing work through Kanban. The "don't do the work yourself" rule and the basic lifecycle are auto-injected into every kanban worker's system prompt; this skill is the deeper playbook…
Orchestrate Claude Code CLI from Hermes — tmux interactive + agent team (stability-first). Print mode is secondary and only used for connectivity smoke tests. Triggers: claude code, cc, delegate to claude, use claude, let claude handle, 用claude, 让claude,…
Source text: Chinese
Use when a Hermes Kanban worker wants to run Codex CLI as an isolated implementation lane while Hermes keeps ownership of task lifecycle, reconciliation, testing, and handoff.
Use when the user wants to create, configure, or migrate a dedicated cron-worker Hermes profile for scheduled/background tasks. Covers profile creation with symlink-based identity sharing (SOUL.md, memory), skill external_dirs configuration, model downgrade…
Orchestrate Claude Code agent teams via `cmux claude-teams` — cmux-native, event-driven (stability-first). cmux 版的 claude-code 编排手册:用 cmux 的 events 事件流 + Feed 决策面板 + workspace 隔离, 取代 raw-tmux 的盲轮询 + PTY 对话框导航。 Triggers: cccmux, cmux claude-teams, cmux 起 team,…
Source text: Chinese
Operate OpenWrt/iStoreOS routers over SSH/LuCI: connect, diagnose network/system state, read UCI config, manage opkg packages cautiously, back up before changes, and avoid lockout. Use when the user asks to connect to, inspect, configure, troubleshoot, or…
Grills a plan or design against the Hermes multi-profile domain model — challenges against CONTEXT.md glossary, cross-references with code and configs, stress-tests with concrete scenarios, and updates documentation inline as decisions crystallise. Structured…
Decrypt DingTalk's local encrypted SQLite database to read group messages, then set up automated polling and delivery. No admin permissions, no developer account needed. Use when the user wants AI to read/monitor DingTalk group chat messages (e.g., class…
Source text: Chinese
Use when producing the daily morning news briefing — multi-source parallel search via web-research-router, fused analysis format (前提→推理→结论 + 趋势 + 为什么重要), mobile + A4 PDF delivery, and TTS voice edition. Supports dual execution mode: Cron (shell-parallel…
Source text: Mixed languages
Set up, configure, and manage Supermemory as Hermes Agent's external memory provider. Covers SDK setup, API key config, provider switching, container_tag isolation for multi-profile deployments, metadata taxonomy, cross-pool wrapper usage, LRU cache layer,…
Use when drafting a technical support email to a SaaS/API vendor about a service outage, data loss, or bug. Covers the full workflow: deep-dive investigation → config audit against official docs → multi-angle testing → evidence gathering → tone-calibrated…
Use when the user wants AI-assisted stock analysis, TradingAgents setup, or to run TauricResearch/TradingAgents locally. Guides installation, API keys, non-interactive runs, report summarization, and risk disclaimers.
Use when managing, evaluating, configuring, or testing text-to-speech providers for Hermes or local agent workflows. Covers provider registry, fallback policy, voice/sample tests, resource benchmarks, artifact/noise checks, and keeping TTS decisions…