skill-forge-plan
starsforks
How to install skill-forge-plan?
agricidaniel/skill-forge/skill-forge-plan
$npx -y skills add AgriciDaniel/skill-forge --skill skill-forge-planArchitecture and design planning for new Claude Code skills. Guides through use case definition, complexity tier selection, sub-skill decomposition, and file structure planning. Use when user says "plan skill", "design skill", "skill architecture", or "skill planning".
SKILL.md
| 1 | # Skill Architecture & Design Planning |
| 2 | |
| 3 | ## Process |
| 4 | |
| 5 | ### Step 1: Domain Discovery |
| 6 | |
| 7 | Ask the user these questions (adapt based on context): |
| 8 | |
| 9 | 1. **What domain is this skill for?** (e.g., SEO, advertising, DevOps, data analysis) |
| 10 | 2. **What are the top 2-3 use cases?** What should users be able to accomplish? |
| 11 | 3. **What trigger phrases would users say?** List 5-10 natural language triggers. |
| 12 | 4. **Does it need external tools?** MCP servers, APIs, CLI tools? |
| 13 | 5. **Who is the target user?** Developer, marketer, analyst, general user? |
| 14 | |
| 15 | ### Step 2: Use Case Decomposition |
| 16 | |
| 17 | For each use case, define: |
| 18 | |
| 19 | ``` |
| 20 | Use Case: [Name] |
| 21 | Trigger: User says "[phrases]" |
| 22 | Steps: |
| 23 | 1. [First action] |
| 24 | 2. [Decision point or validation] |
| 25 | 3. [Next action] |
| 26 | Result: [What success looks like] |
| 27 | Tools Needed: [built-in, MCP, scripts] |
| 28 | ``` |
| 29 | |
| 30 | ### Step 3: Complexity Tier Assessment |
| 31 | |
| 32 | Evaluate based on answers: |
| 33 | |
| 34 | | Signal | Tier 1 | Tier 2 | Tier 3 | Tier 4 | |
| 35 | |--------|--------|--------|--------|--------| |
| 36 | | Use cases | 1-2 | 2-3 | 4-8 | 8+ | |
| 37 | | Needs scripts? | No | Yes | Maybe | Yes | |
| 38 | | Sub-skills needed? | No | No | Yes | Yes | |
| 39 | | Parallel execution? | No | No | No | Yes | |
| 40 | | Reference docs? | No | Maybe | Yes | Yes | |
| 41 | | Industry templates? | No | No | Maybe | Yes | |
| 42 | |
| 43 | **Decision matrix:** |
| 44 | - Single workflow, no scripts -> **Tier 1** (minimal) |
| 45 | - Needs deterministic validation -> **Tier 2** (workflow) |
| 46 | - Multiple distinct workflows -> **Tier 3** (multi-skill) |
| 47 | - Complex domain with parallel delegation -> **Tier 4** (ecosystem) |
| 48 | |
| 49 | ### Step 4: Architecture Design |
| 50 | |
| 51 | Based on tier, generate the architecture: |
| 52 | |
| 53 | **Tier 1 Output:** |
| 54 | ``` |
| 55 | skill-name/ |
| 56 | SKILL.md |
| 57 | ``` |
| 58 | |
| 59 | **Tier 2 Output:** |
| 60 | ``` |
| 61 | skill-name/ |
| 62 | SKILL.md |
| 63 | scripts/ |
| 64 | validate.py |
| 65 | process.py |
| 66 | references/ |
| 67 | domain-knowledge.md |
| 68 | ``` |
| 69 | |
| 70 | **Tier 3 Output:** |
| 71 | ``` |
| 72 | skill-name/ # Main orchestrator |
| 73 | SKILL.md |
| 74 | references/ |
| 75 | shared-reference.md |
| 76 | skills/ |
| 77 | skill-name-sub1/ |
| 78 | SKILL.md |
| 79 | skill-name-sub2/ |
| 80 | SKILL.md |
| 81 | ``` |
| 82 | |
| 83 | **Tier 4 Output:** |
| 84 | ``` |
| 85 | skill-name/ # Main orchestrator |
| 86 | SKILL.md |
| 87 | references/ |
| 88 | ref1.md |
| 89 | ref2.md |
| 90 | scripts/ |
| 91 | script1.py |
| 92 | script2.py |
| 93 | assets/ |
| 94 | template1.md |
| 95 | template2.md |
| 96 | skills/ |
| 97 | skill-name-sub1/ |
| 98 | SKILL.md |
| 99 | skill-name-sub2/ |
| 100 | SKILL.md |
| 101 | ... |
| 102 | agents/ |
| 103 | skill-name-role1.md |
| 104 | skill-name-role2.md |
| 105 | ``` |
| 106 | |
| 107 | ### Step 5: Sub-Skill Decomposition (Tier 3-4 only) |
| 108 | |
| 109 | For each sub-skill, define: |
| 110 | - **Name**: `{parent}-{function}` (kebab-case) |
| 111 | - **Responsibility**: Single, clear purpose |
| 112 | - **Inputs**: What it needs from the orchestrator |
| 113 | - **Outputs**: What it returns |
| 114 | - **Cross-references**: Other sub-skills or references it needs |
| 115 | - **Self-contained?**: Can it run independently or needs orchestration? |
| 116 | |
| 117 | ### Step 6: Routing Table |
| 118 | |
| 119 | Design the command routing: |
| 120 | |
| 121 | ```markdown |
| 122 | | Command | Routes to | Purpose | |
| 123 | |---------|-----------|---------| |
| 124 | | /skill-name | main SKILL.md | Interactive mode | |
| 125 | | /skill-name sub1 | skills/skill-name-sub1/ | Sub-workflow 1 | |
| 126 | | /skill-name sub2 | skills/skill-name-sub2/ | Sub-workflow 2 | |
| 127 | ``` |
| 128 | |
| 129 | ### Step 7: Reference File Planning |
| 130 | |
| 131 | Identify knowledge that should be extracted to reference files: |
| 132 | - Domain-specific rules and thresholds |
| 133 | - Industry templates |
| 134 | - API documentation |
| 135 | - Quality gates and validation criteria |
| 136 | |
| 137 | Rule of thumb: If information is >50 lines and only needed for specific sub-workflows, |
| 138 | extract it to `references/`. |
| 139 | |
| 140 | ### Step 8: Generate Plan Document |
| 141 | |
| 142 | Create a structured plan document: |
| 143 | |
| 144 | ```markdown |
| 145 | # Skill Plan: [name] |
| 146 | |
| 147 | ## Overview |
| 148 | - Domain: [domain] |
| 149 | - Tier: [1-4] |
| 150 | - Sub-skills: [count] |
| 151 | - Scripts: [count] |
| 152 | |
| 153 | ## Use Cases |
| 154 | [list from Step 2] |
| 155 | |
| 156 | ## Architecture |
| 157 | [diagram from Step 4] |
| 158 | |
| 159 | ## Sub-Skills |
| 160 | [details from Step 5] |
| 161 | |
| 162 | ## Routing |
| 163 | [table from Step 6] |
| 164 | |
| 165 | ## Reference Files |
| 166 | [list from Step 7] |
| 167 | |
| 168 | ## Next Steps |
| 169 | Run `/skill-forge build [name]` to scaffold the skill. |
| 170 | ``` |
| 171 | |
| 172 | ## Examples |
| 173 | |
| 174 | ### Example: Planning a DevOps Skill |
| 175 | |
| 176 | User: "I want to create a skill for managing Docker containers and Kubernetes deployments" |
| 177 | |
| 178 | Discovery reveals: |
| 179 | - 6 use cases (container management, K8s deploy, monitoring, logs, scaling, troubleshooting) |
| 180 | - Needs scripts for kubectl and docker commands |
| 181 | - Multiple distinct workflows |
| 182 | - Cross-references between monitoring and troubleshooting |
| 183 | |
| 184 | Assessment: **Tier 3** (multi-skill orchestrator) |
| 185 | |
| 186 | Architecture: |
| 187 | ``` |
| 188 | devops/ # Main orchestrator |
| 189 | SKILL.md |
| 190 | scripts/ |
| 191 | health_check.py # Cluster health check |
| 192 | log_parser.py # Log analysis |
| 193 | references/ |
| 194 | k8s-patterns.md # Deployment patterns |
| 195 | docker-best-practices.md |
| 196 | skills/ |
| 197 | devops-docker/SKILL.md # Container management |
| 198 | devops-k8s/SKILL.md # Kubernetes deployments |
| 199 | devops-monitor/SKILL.md # Monitoring and alerts |
| 200 | devops-logs/SKILL.md # L |
People who liked this also viewed
6 picksType
- lark-workflow-meeting-summary会议纪要整理工作流：汇总指定时间范围内的会议纪要并生成结构化报告。当用户需要整理会议纪要、生成会议周报、回顾一段时间内的会议内容时使用。SkillsJul 2026
- lark-workflow-standup-report日程待办摘要：编排 calendar +agenda 和 task +get-my-tasks，生成指定日期的日程与未完成任务摘要。适用于了解今天/明天/本周的安排。SkillsJul 2026
- lark-approval飞书审批：查询和处理审批待办/已办/实例，搜索可发起审批定义、查看定义详情并发起原生审批实例。当用户要处理审批任务、查看审批实例、搜索或发起审批时使用。审批待办不是飞书任务；非审批类待办走 lark-task。不负责创建审批定义；三方审批定义不走原生提单。SkillsJul 2026
- lark-okr飞书 OKR：管理目标与关键结果。查看和编辑 OKR 周期、目标、关键结果、对齐关系、量化指标和进展记录。当用户需要查看或创建 OKR、管理目标和关键结果、查看对齐关系时使用。不负责：待办任务管理（lark-task）、日程/会议安排（lark-calendar）、绩效评估SkillsJul 2026
- brainstormingYou MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.SkillsJul 2026
- simpleInvoke before any creative or architectural work — feature design, component creation, or behavioral changes. A streamlined brainstorming process optimized for fast, focused decision-making. Gen AgentSkillsMar 2026
