[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AYVW5SWWEAWXPA4AMS3XRZ":3,"skill-same:01M7AYVW5SWWEAWXPA4AMS3XRZ":62},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":33,"commitSha":34,"skillDirectory":35,"forksCount":36,"repositoryHtmlUrl":37,"repositoryHomepageUrl":6,"ownerHtmlUrl":38,"ownerBlogUrl":39,"ownerDescription":40,"updatedAt":41,"install":42,"resources":55},"01M7AYVW5SWWEAWXPA4AMS3XRZ","spec-document-elicitation",null,"Elicit requirements from stakeholders via structured 'grill me' interviews and populate spec documents from templates. Covers the full 01_requirement phase.","oat431\u002Fhermes_config_backup","oat431","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54722120?v=4","01M7AYD07EH8SYPMYY8CGZKS22","Oralita","01M7AYD0816QDF77E3TWCRSQC9",1,"Unlicense","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fproduct-owner\u002Fspec-document-elicitation\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: spec-document-elicitation\ndescription: \"Elicit requirements from stakeholders via structured 'grill me' interviews and populate spec documents from templates. Covers the full 01_requirement phase.\"\ntags: [requirements, elicitation, product-owner, ba, interview, templates, specification, babok]\n---\n\n# Spec Document Elicitation\n\nElicit requirements from a stakeholder (the user) through structured Q&A and populate standardized specification documents from templates. The \"grill me\" pattern: sequential `clarify()` calls with curated multiple-choice options, each answer feeding the next question and the document being written.\n\n## Trigger Conditions\n\nLoad this skill when:\n- A user asks to \"grill me,\" \"interview me,\" or \"ask me questions\" to set up project specifications\n- A user references the `project_spec` repository (`F:\\projects\\project_spec`) and wants foundation documents populated\n- A user wants Business Objectives, User Stories, Acceptance Criteria, or Stakeholder Analysis written from scratch\n- A user says \"I'm ready, ask me\" in the context of a new project\u002Fspec\n\n## Repository Structure\n\n```\nF:\\projects\\project_spec\\\n├── README.md                          # Platform overview & service catalog\n├── template\\                          # Blank templates (the source of truth)\n│   └── 01_requirement\\\n│       ├── 011_business_objective.md  # SMART goals, KPIs, risk matrix\n│       ├── 012_user_stories.md        # As-a-I-want-so-that, epics, story maps\n│       ├── 013_acceptance_criteria.md # Given-When-Then BDD format\n│       ├── 014_stakeholder_analysis.md # Influence\u002Finterest matrix\n│       └── 015_definition_of_done.md  # DoD criteria\n│   └── 02_design\\ ... 07_pm\\         # Later phases\n└── spec\\                              # Populated specs (one dir per project)\n    └── \u003Cproject_name>\\\n        └── 01_requirement\\            # Where documents get written\n```\n\n**Reference example**: `spec\u002Ftiny_mchwa\u002F` — a completed spec to study for format and depth.\n\n## Elicitation Workflow\n\n### Phase 1: Discovery (read, don't ask)\n1. Read `F:\\projects\\project_spec\\README.md` — understand the project context\n2. Read all relevant templates from `F:\\projects\\project_spec\\template\\01_requirement\\`\n3. Read an existing completed spec (e.g., `spec\u002Ftiny_mchwa\u002F`) as a reference\n4. Scan the project's spec directory to see what already exists\n\n### Phase 1.5: Structure Decision (multi-service projects only)\n\n**When the project has multiple discrete services** (e.g., a platform with foundation services + business microservices), ask the structure question BEFORE writing any document beyond the platform-level Business Objectives:\n\n> \"Should each service get its own deep-dive documents, with the platform serving as an umbrella? Or should all stories live in one monolithic platform doc?\"\n\n**The recommended pattern (and the one to lead with):**\n\n```\nspec\u002F\n├── \u003Cplatform>\u002F                     ← UMBRELLA: platform-level docs only\n│   ├── README.md                   ← Architecture overview, service map, quick links\n│   └── 01_requirement\u002F\n│       ├── 011_business_objective.md   ← Cross-cutting objectives (all services)\n│       ├── 012_user_stories.md         ← Consolidated story map + links to service docs\n│       └── 014_stakeholder_analysis.md ← All personas (human + AI)\n│\n├── \u003Cservice-a>\u002F                    ← DEEP DIVE: self-contained, AI-persona-ready\n│   └── 01_requirement\u002F\n│       ├── 011_business_objective.md   ← Service-specific objectives (subset of platform)\n│       ├── 012_user_stories.md         ← That service's stories only\n│       └── 013_acceptance_criteria.md  ← BDD criteria for that service's stories\n│\n├── \u003Cservice-b>\u002F                    ← Same structure, fully independent\n│   └── 01_requirement\u002F\n│       └── ...\n```\n\n**Why this pattern:**\n- Each service spec is self-contained — an AI persona can pick up one directory and build without reading the whole platform\n- The platform umbrella provides cross-service traceability (story → objective, service → dependency) without duplicating content\n- Story points, epics, and milestones are consolidated at the platform level; details live in service docs\n- `012_user_stories.md` at the platform level is a **thin umbrella** — story map + cross-service dependencies + links. Not a copy of service-level stories.\n\n**When NOT to split:**\n- Single-service projects (e.g., one microservice with no platform layer)\n- The platform document is the only document needed (no deep dives)\n- Tiny projects where splitting adds more overhead than value\n\n### Phase 2: Grilling (structured Q&A)\nUse `clarify()` with multiple-choice options. Sequencing matters — start broad, then narrow:\n\n**Round 1 — The Big Picture:**\n- Primary motivation (why build this?)\n- Users and scale (self only, friends, public-ready)\n- Timeline and delivery preferences (docs first? code first?)\n\n**Round 2 — Scope & Architecture:**\n- MVP scope (what's the minimum viable set?)\n- Technology preferences or constraints\n- Deployment environment (Docker Compose, Kubernetes, bare metal)\n\n**Round 3 — Drilling into specifics:**\n- For each component\u002Fservice in scope, ask targeted questions\n- Use prior answers to constrain subsequent choices\n- Each answer should narrow options for the next question\n\n**Rules for grill questions:**\n- ALWAYS provide 3-4 curated choices via `clarify(choices=[...])` — never open-ended unless no good options exist\n- Design choices that progressively narrow scope\n- Each choice must force a real decision between distinct paths\n- After 4-5 rounds of questions, you should have enough to write a document\n- If the user picks \"Other,\" honor it and adapt\n\n### Phase 3: Document Population\nFor each document, follow the template structure precisely:\n\n1. Copy the template's frontmatter — replace placeholders with real values\n2. Fill every section — empty sections with placeholder text are unacceptable\n3. Use real data from the interview — never fabricate metrics, names, or numbers\n4. Mark unknowns honestly — use \"TBD\" rather than guessing\n5. Trace stories to objectives — each user story must map to a business objective\n6. Include \"Related Documents\" cross-references at the bottom\n\n### Document Order (dependencies)\n\n**For the platform umbrella:**\n1. `011_business_objective.md` — defines the WHY (objectives, KPIs, success criteria) for all services\n2. `012_user_stories.md` — consolidated story map + inter-service traceability + links (thin umbrella, NOT a copy)\n3. `014_stakeholder_analysis.md` — defines WHO CARES (influence, concerns, engagement)\n\n**For each service deep-dive:**\n1. `011_business_objective.md` — service-specific objectives (subset of platform objectives)\n2. `012_user_stories.md` — that service's stories only, self-contained for an AI persona\n3. `013_acceptance_criteria.md` — BDD criteria for that service's stories (Given-When-Then)\n\n**Rule:** The platform umbrella's `012_user_stories.md` links to service-level story docs. It does NOT duplicate them. Each service doc is the source of truth for its own stories.\n\n### Phase 4: Present & Continue\nAfter each document:\n- Summarize what was written (2-3 lines)\n- Show key metrics (e.g., \"5 objectives,\" \"18 stories across 4 epics\")\n- Ask \"Review or continue?\" — don't assume all documents in one session\n\n## Diagram Preferences\n\n- **Architecture \u002F data flow diagrams**: `flowchart TB` or `flowchart LR` with `subgraph` blocks, solid arrows (`-->`) for data flow, dashed (`-.->`) for cross-cutting concerns (auth validation, metrics scraping, service registration). Color-code nodes by layer (edge, foundation, gateway, observability, business).\n- **Project \u002F directory structure**: `treeView-beta` with `├──` \u002F `└──` tree lines and `→` annotations for descriptions. NEVER use `mindmap` for project structure — the user explicitly corrected this. NEVER use ASCII art boxes.\n- **Dependency graphs**: `flowchart` with color-coded priority nodes (🔴 red fill for must-have, 🟡 orange for should-have, 🟢 green for nice-to-have).\n- **Stakeholder maps**: `quadrantChart` with labeled quadrants.\n- **Execution order \u002F sprint plans**: `flowchart LR` with colored initiative nodes and dependency arrows.\n- Always wrap Mermaid in ```mermaid fences. Mermaid renders natively in Obsidian and GitHub.\n\n## Design Review Feedback Loop\n\n> Requirements documents are NOT static. After PO writes `01_requirement`, the Design\u002FDev persona reviews them against real architecture constraints and produces a **meeting minute** (placed in `spec\u002Fmeeting-minute\u002F`) listing decisions and affected documents. The PO must then update ALL affected docs and produce a **response meeting minute** confirming the changes.\n\n### The feedback loop pattern:\n\n```\nPO writes 01_requirement docs\n  → Design\u002FDev persona reviews against real infra\n  → Design\u002FDev produces meeting minute (affect-doc.md) with:\n     - Decisions made (D1, D2, ... Dn)\n     - Documents requiring changes (severity: 🔴 Critical \u002F 🟡 Moderate)\n     - Specific line-level changes needed\n  → PO reads the meeting minute\n  → PO grills user on any unclear decisions\n  → PO updates ALL listed documents (bump version to 0.2)\n  → PO writes a response meeting minute confirming:\n     - What changed in each document\n     - Updated story\u002Fpoint\u002FAC counts\n     - Architecture rules (the new ground truth)\n  → PO hands back to Design persona\n```\n\n### Key rules for the feedback loop:\n- **Read the meeting minute completely** before touching any document. Understand the full scope of changes.\n- **Grill the user on ambiguous decisions** — some design decisions need PO clarification (e.g., \"does Gate see auth traffic at all?\" or \"is Valkey for rate limiting or something else?\").\n- **Update docs in order of severity** — 🔴 Critical docs first (they block development).\n- **Don't forget the README** — both platform-level and root-level READMEs need updating when architecture changes. This is easy to miss.\n- **Bump document versions** — changed docs go from `0.1` → `0.2` with a revision history entry.\n- **Produce a response meeting minute** — this is the contract that tells the next persona what changed and that requirements are now aligned.\n\n## Post-Requirements Deliverables\n\nAfter `01_requirement` is complete and services are being built\u002Fdeployed, the PO persona also produces:\n\n### Phase Plans (`plan\u002F` directory)\nWhen the DevOps\u002FDev persona proposes a new phase (e.g., \"Phase 2: add CI\u002FCD, observability, alerting\"), the PO:\n1. Reads the DevOps proposal meeting minute\n2. Grills the user on decisions (scope, notification channel, deploy strategy, retention policy)\n3. Writes a structured Phase Plan with: initiatives, PO decisions, execution order (Mermaid), action items, Definition of Done, and transition criteria\n\n### Construction Overview (`03_construction\u002F031_README_developer_guide.md`)\nA platform-level document that ties together each service's `03_construction\u002F` folder. Contains:\n- Service inventory (treeView-beta diagrams)\n- Monorepo structure\n- Development workflow (local dev + CI\u002FCD deploy flow)\n- Build profiles per tech stack\n- Shared resources table\n- Coding standards summary\n\n### Meeting Minutes (`spec\u002Fmeeting-minute\u002F`)\nEvery cross-persona handoff produces a meeting minute. These are the **contracts** between personas:\n- Format: YAML frontmatter + structured sections (decisions, documents affected, action items)\n- Named: `\u003Ctopic>_\u003Cdate>.md` (e.g., `po-update-2026-07-22.md`)\n- Include severity tags (🔴\u002F🟡) on affected documents\n- Include a handoff checklist at the end\n\n## Pitfalls\n\n- **DON'T write all documents before the user has seen any.** Write one, present it, get implicit or explicit go-ahead, then write the next.\n- **DON'T ask the same question twice.** Track what's been answered. Each round builds on prior answers.\n- **DON'T use open-ended `clarify()` when you can provide choices.** The user chose \"grill me\" mode — they want to pick from options.\n- **DON'T fill template fields with placeholder text.** If you lack data, ask another question or mark it `TBD`.\n- **DON'T skip sections assuming they're irrelevant for a small project.** A portfolio project needs full documents. Every section signals rigor.\n- **When the user changes a fundamental assumption mid-session** (e.g., \"Guard IS the Keycloak instance itself\"), note that earlier documents may need updating. Flag it explicitly.\n- **DON'T write monolithic `012_user_stories.md` with all services' stories in one file for multi-service platforms.** This forces every persona to read everything. Ask the structure question (Phase 1.5) BEFORE populating story documents. If the user wants service-level docs, split the stories and make the platform `012` a thin umbrella.\n- **DON'T restructure mid-session without asking.** If the user suggests a structural change, pause and confirm the pattern before applying it everywhere. \"Are all three foundation services getting independent docs? What about business services?\"\n- **DON'T forget the READMEs when architecture changes.** When applying design review changes, both the platform-level README (`spec\u002F\u003Cplatform>\u002FREADME.md`) AND the root-level README (`project_spec\u002FREADME.md`) must be updated with new architecture diagrams, service statuses, ports, domains, and phase tracking. These are high-visibility documents — stale information here undermines trust in the rest of the spec.\n- **DON'T replace entire table rows when adding rows.** Use targeted `patch` operations with enough context to be unique. Replacing a row's content when you meant to insert a new row silently deletes data.\n","# Spec Document Elicitation\n\nElicit requirements from a stakeholder (the user) through structured Q&A and populate standardized specification documents from templates. The \"grill me\" pattern: sequential `clarify()` calls with curated multiple-choice options, each answer feeding the next question and the document being written.\n\n## Trigger Conditions\n\nLoad this skill when:\n- A user asks to \"grill me,\" \"interview me,\" or \"ask me questions\" to set up project specifications\n- A user references the `project_spec` repository (`F:\\projects\\project_spec`) and wants foundation documents populated\n- A user wants Business Objectives, User Stories, Acceptance Criteria, or Stakeholder Analysis written from scratch\n- A user says \"I'm ready, ask me\" in the context of a new project\u002Fspec\n\n## Repository Structure\n\n```\nF:\\projects\\project_spec\\\n├── README.md                          # Platform overview & service catalog\n├── template\\                          # Blank templates (the source of truth)\n│   └── 01_requirement\\\n│       ├── 011_business_objective.md  # SMART goals, KPIs, risk matrix\n│       ├── 012_user_stories.md        # As-a-I-want-so-that, epics, story maps\n│       ├── 013_acceptance_criteria.md # Given-When-Then BDD format\n│       ├── 014_stakeholder_analysis.md # Influence\u002Finterest matrix\n│       └── 015_definition_of_done.md  # DoD criteria\n│   └── 02_design\\ ... 07_pm\\         # Later phases\n└── spec\\                              # Populated specs (one dir per project)\n    └── \u003Cproject_name>\\\n        └── 01_requirement\\            # Where documents get written\n```\n\n**Reference example**: `spec\u002Ftiny_mchwa\u002F` — a completed spec to study for format and depth.\n\n## Elicitation Workflow\n\n### Phase 1: Discovery (read, don't ask)\n1. Read `F:\\projects\\project_spec\\README.md` — understand the project context\n2. Read all relevant templates from `F:\\projects\\project_spec\\template\\01_requirement\\`\n3. Read an existing completed spec (e.g., `spec\u002Ftiny_mchwa\u002F`) as a reference\n4. Scan the project's spec directory to see what already exists\n\n### Phase 1.5: Structure Decision (multi-service projects only)\n\n**When the project has multiple discrete services** (e.g., a platform with foundation services + business microservices), ask the structure question BEFORE writing any document beyond the platform-level Business Objectives:\n\n> \"Should each service get its own deep-dive documents, with the platform serving as an umbrella? Or should all stories live in one monolithic platform doc?\"\n\n**The recommended pattern (and the one to lead with):**\n\n```\nspec\u002F\n├── \u003Cplatform>\u002F                     ← UMBRELLA: platform-level docs only\n│   ├── README.md                   ← Architecture overview, service map, quick links\n│   └── 01_requirement\u002F\n│       ├── 011_business_objective.md   ← Cross-cutting objectives (all services)\n│       ├── 012_user_stories.md         ← Consolidated story map + links to service docs\n│       └── 014_stakeholder_analysis.md ← All personas (human + AI)\n│\n├── \u003Cservice-a>\u002F                    ← DEEP DIVE: self-contained, AI-persona-ready\n│   └── 01_requirement\u002F\n│       ├── 011_business_objective.md   ← Service-specific objectives (subset of platform)\n│       ├── 012_user_stories.md         ← That service's stories only\n│       └── 013_acceptance_criteria.md  ← BDD criteria for that service's stories\n│\n├── \u003Cservice-b>\u002F                    ← Same structure, fully independent\n│   └── 01_requirement\u002F\n│       └── ...\n```\n\n**Why this pattern:**\n- Each service spec is self-contained — an AI persona can pick up one directory and build without reading the whole platform\n- The platform umbrella provides cross-service traceability (story → objective, service → dependency) without duplicating content\n- Story points, epics, and milestones are consolidated at the platform level; details live in service docs\n- `012_user_stories.md` at the platform level is a **thin umbrella** — story map + cross-service dependencies + links. Not a copy of service-level stories.\n\n**When NOT to split:**\n- Single-service projects (e.g., one microservice with no platform layer)\n- The platform document is the only document needed (no deep dives)\n- Tiny projects where splitting adds more overhead than value\n\n### Phase 2: Grilling (structured Q&A)\nUse `clarify()` with multiple-choice options. Sequencing matters — start broad, then narrow:\n\n**Round 1 — The Big Picture:**\n- Primary motivation (why build this?)\n- Users and scale (self only, friends, public-ready)\n- Timeline and delivery preferences (docs first? code first?)\n\n**Round 2 — Scope & Architecture:**\n- MVP scope (what's the minimum viable set?)\n- Technology preferences or constraints\n- Deployment environment (Docker Compose, Kubernetes, bare metal)\n\n**Round 3 — Drilling into specifics:**\n- For each component\u002Fservice in scope, ask targeted questions\n- Use prior answers to constrain subsequent choices\n- Each answer should narrow options for the next question\n\n**Rules for grill questions:**\n- ALWAYS provide 3-4 curated choices via `clarify(choices=[...])` — never open-ended unless no good options exist\n- Design choices that progressively narrow scope\n- Each choice must force a real decision between distinct paths\n- After 4-5 rounds of questions, you should have enough to write a document\n- If the user picks \"Other,\" honor it and adapt\n\n### Phase 3: Document Population\nFor each document, follow the template structure precisely:\n\n1. Copy the template's frontmatter — replace placeholders with real values\n2. Fill every section — empty sections with placeholder text are unacceptable\n3. Use real data from the interview — never fabricate metrics, names, or numbers\n4. Mark unknowns honestly — use \"TBD\" rather than guessing\n5. Trace stories to objectives — each user story must map to a business objective\n6. Include \"Related Documents\" cross-references at the bottom\n\n### Document Order (dependencies)\n\n**For the platform umbrella:**\n1. `011_business_objective.md` — defines the WHY (objectives, KPIs, success criteria) for all services\n2. `012_user_stories.md` — consolidated story map + inter-service traceability + links (thin umbrella, NOT a copy)\n3. `014_stakeholder_analysis.md` — defines WHO CARES (influence, concerns, engagement)\n\n**For each service deep-dive:**\n1. `011_business_objective.md` — service-specific objectives (subset of platform objectives)\n2. `012_user_stories.md` — that service's stories only, self-contained for an AI persona\n3. `013_acceptance_criteria.md` — BDD criteria for that service's stories (Given-When-Then)\n\n**Rule:** The platform umbrella's `012_user_stories.md` links to service-level story docs. It does NOT duplicate them. Each service doc is the source of truth for its own stories.\n\n### Phase 4: Present & Continue\nAfter each document:\n- Summarize what was written (2-3 lines)\n- Show key metrics (e.g., \"5 objectives,\" \"18 stories across 4 epics\")\n- Ask \"Review or continue?\" — don't assume all documents in one session\n\n## Diagram Preferences\n\n- **Architecture \u002F data flow diagrams**: `flowchart TB` or `flowchart LR` with `subgraph` blocks, solid arrows (`-->`) for data flow, dashed (`-.->`) for cross-cutting concerns (auth validation, metrics scraping, service registration). Color-code nodes by layer (edge, foundation, gateway, observability, business).\n- **Project \u002F directory structure**: `treeView-beta` with `├──` \u002F `└──` tree lines and `→` annotations for descriptions. NEVER use `mindmap` for project structure — the user explicitly corrected this. NEVER use ASCII art boxes.\n- **Dependency graphs**: `flowchart` with color-coded priority nodes (🔴 red fill for must-have, 🟡 orange for should-have, 🟢 green for nice-to-have).\n- **Stakeholder maps**: `quadrantChart` with labeled quadrants.\n- **Execution order \u002F sprint plans**: `flowchart LR` with colored initiative nodes and dependency arrows.\n- Always wrap Mermaid in ```mermaid fences. Mermaid renders natively in Obsidian and GitHub.\n\n## Design Review Feedback Loop\n\n> Requirements documents are NOT static. After PO writes `01_requirement`, the Design\u002FDev persona reviews them against real architecture constraints and produces a **meeting minute** (placed in `spec\u002Fmeeting-minute\u002F`) listing decisions and affected documents. The PO must then update ALL affected docs and produce a **response meeting minute** confirming the changes.\n\n### The feedback loop pattern:\n\n```\nPO writes 01_requirement docs\n  → Design\u002FDev persona reviews against real infra\n  → Design\u002FDev produces meeting minute (affect-doc.md) with:\n     - Decisions made (D1, D2, ... Dn)\n     - Documents requiring changes (severity: 🔴 Critical \u002F 🟡 Moderate)\n     - Specific line-level changes needed\n  → PO reads the meeting minute\n  → PO grills user on any unclear decisions\n  → PO updates ALL listed documents (bump version to 0.2)\n  → PO writes a response meeting minute confirming:\n     - What changed in each document\n     - Updated story\u002Fpoint\u002FAC counts\n     - Architecture rules (the new ground truth)\n  → PO hands back to Design persona\n```\n\n### Key rules for the feedback loop:\n- **Read the meeting minute completely** before touching any document. Understand the full scope of changes.\n- **Grill the user on ambiguous decisions** — some design decisions need PO clarification (e.g., \"does Gate see auth traffic at all?\" or \"is Valkey for rate limiting or something else?\").\n- **Update docs in order of severity** — 🔴 Critical docs first (they block development).\n- **Don't forget the README** — both platform-level and root-level READMEs need updating when architecture changes. This is easy to miss.\n- **Bump document versions** — changed docs go from `0.1` → `0.2` with a revision history entry.\n- **Produce a response meeting minute** — this is the contract that tells the next persona what changed and that requirements are now aligned.\n\n## Post-Requirements Deliverables\n\nAfter `01_requirement` is complete and services are being built\u002Fdeployed, the PO persona also produces:\n\n### Phase Plans (`plan\u002F` directory)\nWhen the DevOps\u002FDev persona proposes a new phase (e.g., \"Phase 2: add CI\u002FCD, observability, alerting\"), the PO:\n1. Reads the DevOps proposal meeting minute\n2. Grills the user on decisions (scope, notification channel, deploy strategy, retention policy)\n3. Writes a structured Phase Plan with: initiatives, PO decisions, execution order (Mermaid), action items, Definition of Done, and transition criteria\n\n### Construction Overview (`03_construction\u002F031_README_developer_guide.md`)\nA platform-level document that ties together each service's `03_construction\u002F` folder. Contains:\n- Service inventory (treeView-beta diagrams)\n- Monorepo structure\n- Development workflow (local dev + CI\u002FCD deploy flow)\n- Build profiles per tech stack\n- Shared resources table\n- Coding standards summary\n\n### Meeting Minutes (`spec\u002Fmeeting-minute\u002F`)\nEvery cross-persona handoff produces a meeting minute. These are the **contracts** between personas:\n- Format: YAML frontmatter + structured sections (decisions, documents affected, action items)\n- Named: `\u003Ctopic>_\u003Cdate>.md` (e.g., `po-update-2026-07-22.md`)\n- Include severity tags (🔴\u002F🟡) on affected documents\n- Include a handoff checklist at the end\n\n## Pitfalls\n\n- **DON'T write all documents before the user has seen any.** Write one, present it, get implicit or explicit go-ahead, then write the next.\n- **DON'T ask the same question twice.** Track what's been answered. Each round builds on prior answers.\n- **DON'T use open-ended `clarify()` when you can provide choices.** The user chose \"grill me\" mode — they want to pick from options.\n- **DON'T fill template fields with placeholder text.** If you lack data, ask another question or mark it `TBD`.\n- **DON'T skip sections assuming they're irrelevant for a small project.** A portfolio project needs full documents. Every section signals rigor.\n- **When the user changes a fundamental assumption mid-session** (e.g., \"Guard IS the Keycloak instance itself\"), note that earlier documents may need updating. Flag it explicitly.\n- **DON'T write monolithic `012_user_stories.md` with all services' stories in one file for multi-service platforms.** This forces every persona to read everything. Ask the structure question (Phase 1.5) BEFORE populating story documents. If the user wants service-level docs, split the stories and make the platform `012` a thin umbrella.\n- **DON'T restructure mid-session without asking.** If the user suggests a structural change, pause and confirm the pattern before applying it everywhere. \"Are all three foundation services getting independent docs? What about business services?\"\n- **DON'T forget the READMEs when architecture changes.** When applying design review changes, both the platform-level README (`spec\u002F\u003Cplatform>\u002FREADME.md`) AND the root-level README (`project_spec\u002FREADME.md`) must be updated with new architecture diagrams, service statuses, ports, domains, and phase tracking. These are high-visibility documents — stale information here undermines trust in the rest of the spec.\n- **DON'T replace entire table rows when adding rows.** Use targeted `patch` operations with enough context to be unique. Replacing a row's content when you meant to insert a new row silently deletes data.\n",{"name":5,"tags":24,"description":7},[25,26,27,28,29,30,31,32],"requirements","elicitation","product-owner","ba","interview","templates","specification","babok","main","2988869c5a7d7be2225c9ea7a54a4a56a59b0d45","workflow\u002Fproduct-owner\u002Fspec-document-elicitation",0,"https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup","https:\u002F\u002Fgithub.com\u002Foat431","https:\u002F\u002Fpanomete.com\u002F","Hope and Luck is not a part of the strategy, \r\n\r\nSoftskill produces plan Hard skill produce product.","2026-11-03 00:43:08",{"source":8,"skillName":5,"commands":43,"prompt":53,"zipAvailable":54},[44,47,50],{"tool":45,"command":46},"npx","npx skills add oat431\u002Fhermes_config_backup --skill spec-document-elicitation",{"tool":48,"command":49},"bunx","bunx skills add oat431\u002Fhermes_config_backup --skill spec-document-elicitation",{"tool":51,"command":52},"pnpm","pnpm dlx skills add oat431\u002Fhermes_config_backup --skill spec-document-elicitation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fproduct-owner\u002Fspec-document-elicitation\u002FSKILL.md",true,[56],{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Ftemplate-field-guide.md","REFERENCE","参考文档",4519,"https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fproduct-owner\u002Fspec-document-elicitation\u002Freferences\u002Ftemplate-field-guide.md",[63,71,78,85,92,99],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":67,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7AYZ2JY3NFKAPQSC39N5F3F","project-spec-docs","Fill\u002Fupdate project_spec docs (templates, MM handoffs).","MIT","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fui-ux\u002Fproject-spec-docs\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":67,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7AYYRZPE7HQX7778558X52S","penpot-mcp","Penpot MCP: connect, verify, and fix streaming via nginx.","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fui-ux\u002Fpenpot-mcp\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7AYYNKP61CN2281VMJDXMB5","nginx-streaming-proxy","Configure nginx for streaming endpoints (SSE, MCP, WS).","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fui-ux\u002Fnginx-streaming-proxy\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":67,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7AYYJP0MDV1D3WVEEZB3SGA","branding-ci","Use for brand\u002FCI identity, style guides, and brand docs.","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fui-ux\u002Fbranding-ci\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":19},"01M7AYYGCXVT97DDTCM0ZJT6F6","yuanbao","Yuanbao (元宝) groups: @mention users, query info\u002Fmembers.","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fshared-skills\u002Fyuanbao\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":19},"01M7AYYC2Y65NMGMD4GQXAZMMS","slidev","Create and present web-based slidedecks for developers using Slidev with Markdown, Vue components, code highlighting, animations, and interactive features. Use when building technical presentations, conference talks, code walkthroughs, teaching materials, or developer decks.","https:\u002F\u002Fgithub.com\u002Foat431\u002Fhermes_config_backup\u002Fblob\u002Fmain\u002Fworkflow\u002Fshared-skills\u002Fslidev\u002FSKILL.md",[],[]]