[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77JM9FFPAWXF87PPFCYN7JN":3,"skill-same:01M77JM9FFPAWXF87PPFCYN7JN":169},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M77JM9FFPAWXF87PPFCYN7JN","create-plans",null,"Create hierarchical project plans optimized for solo agentic development. Use when planning projects, phases, or tasks that Claude will execute. Produces Claude-executable plans with verification criteria, not enterprise documentation. Handles briefs, roadmaps, phase plans, and context handoffs.","Elad73\u002Fagentic-base-layer","Elad73","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11886649?v=4","01M77JKHHH0GYX9ZBKNJ38QRAJ","EladRon","01M77JKHJNDTJ53993X2QNMHSG",0,"MIT","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: create-plans\ndescription: Create hierarchical project plans optimized for solo agentic development. Use when planning projects, phases, or tasks that Claude will execute. Produces Claude-executable plans with verification criteria, not enterprise documentation. Handles briefs, roadmaps, phase plans, and context handoffs.\n---\n\n\u003Cessential_principles>\n\n\u003Cprinciple name=\"solo_developer_plus_claude\">\nYou are planning for ONE person (the user) and ONE implementer (Claude).\nNo teams. No stakeholders. No ceremonies. No coordination overhead.\nThe user is the visionary\u002Fproduct owner. Claude is the builder.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"plans_are_prompts\">\nPLAN.md is not a document that gets transformed into a prompt.\nPLAN.md IS the prompt. It contains:\n- Objective (what and why)\n- Context (@file references)\n- Tasks (type, files, action, verify, done, checkpoints)\n- Verification (overall checks)\n- Success criteria (measurable)\n- Output (SUMMARY.md specification)\n\nWhen planning a phase, you are writing the prompt that will execute it.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"scope_control\">\nPlans must complete within ~50% of context usage to maintain consistent quality.\n\n**The quality degradation curve:**\n- 0-30% context: Peak quality (comprehensive, thorough, no anxiety)\n- 30-50% context: Good quality (engaged, manageable pressure)\n- 50-70% context: Degrading quality (efficiency mode, compression)\n- 70%+ context: Poor quality (self-lobotomization, rushed work)\n\n**Critical insight:** Claude doesn't degrade at 80% - it degrades at ~40-50% when it sees context mounting and enters \"completion mode.\" By 80%, quality has already crashed.\n\n**Solution:** Aggressive atomicity - split phases into many small, focused plans.\n\nExamples:\n- `01-01-PLAN.md` - Phase 1, Plan 1 (2-3 tasks: database schema only)\n- `01-02-PLAN.md` - Phase 1, Plan 2 (2-3 tasks: database client setup)\n- `01-03-PLAN.md` - Phase 1, Plan 3 (2-3 tasks: API routes)\n- `01-04-PLAN.md` - Phase 1, Plan 4 (2-3 tasks: UI components)\n\nEach plan is independently executable, verifiable, and scoped to **2-3 tasks maximum**.\n\n**Atomic task principle:** Better to have 10 small, high-quality plans than 3 large, degraded plans. Each commit should be surgical, focused, and maintainable.\n\n**Autonomous execution:** Plans without checkpoints execute via subagent with fresh context - impossible to degrade.\n\nSee: references\u002Fscope-estimation.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"human_checkpoints\">\n**Claude automates everything that has a CLI or API.** Checkpoints are for verification and decisions, not manual work.\n\n**Checkpoint types:**\n- `checkpoint:human-verify` - Human confirms Claude's automated work (visual checks, UI verification)\n- `checkpoint:decision` - Human makes implementation choice (auth provider, architecture)\n\n**Rarely needed:** `checkpoint:human-action` - Only for actions with no CLI\u002FAPI (email verification links, account approvals requiring web login with 2FA)\n\n**Critical rule:** If Claude CAN do it via CLI\u002FAPI\u002Ftool, Claude MUST do it. Never ask human to:\n- Deploy to Vercel\u002FRailway\u002FFly (use CLI)\n- Create Stripe webhooks (use CLI\u002FAPI)\n- Run builds\u002Ftests (use Bash)\n- Write .env files (use Write tool)\n- Create database resources (use provider CLI)\n\n**Protocol:** Claude automates work → reaches checkpoint:human-verify → presents what was done → waits for confirmation → resumes\n\nSee: references\u002Fcheckpoints.md, references\u002Fcli-automation.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"deviation_rules\">\nPlans are guides, not straitjackets. Real development always involves discoveries.\n\n**During execution, deviations are handled automatically via 5 embedded rules:**\n\n1. **Auto-fix bugs** - Broken behavior → fix immediately, document in Summary\n2. **Auto-add missing critical** - Security\u002Fcorrectness gaps → add immediately, document\n3. **Auto-fix blockers** - Can't proceed → fix immediately, document\n4. **Ask about architectural** - Major structural changes → stop and ask user\n5. **Log enhancements** - Nice-to-haves → auto-log to ISSUES.md, continue\n\n**No user intervention needed for Rules 1-3, 5.** Only Rule 4 (architectural) requires user decision.\n\n**All deviations documented in Summary** with: what was found, what rule applied, what was done, commit hash.\n\n**Result:** Flow never breaks. Bugs get fixed. Scope stays controlled. Complete transparency.\n\nSee: workflows\u002Fexecute-phase.md (deviation_rules section)\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"ship_fast_iterate_fast\">\nNo enterprise process. No approval gates. No multi-week timelines.\nPlan → Execute → Ship → Learn → Repeat.\n\n**Milestone-driven:** Ship v1.0 → mark milestone → plan v1.1 → ship → repeat.\nMilestones mark shipped versions and enable continuous iteration.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"milestone_boundaries\">\nMilestones mark shipped versions (v1.0, v1.1, v2.0).\n\n**Purpose:**\n- Historical record in MILESTONES.md (what shipped when)\n- Greenfield → Brownfield transition marker\n- Git tags for releases\n- Clear completion rituals\n\n**Default approach:** Extend existing roadmap with new phases.\n- v1.0 ships (phases 1-4) → add phases 5-6 for v1.1\n- Continuous phase numbering (01-99)\n- Milestone groupings keep roadmap organized\n\n**Archive ONLY for:** Separate codebases or complete rewrites (rare).\n\nSee: references\u002Fmilestone-management.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"anti_enterprise_patterns\">\nNEVER include in plans:\n- Team structures, roles, RACI matrices\n- Stakeholder management, alignment meetings\n- Sprint ceremonies, standups, retros\n- Multi-week estimates, resource allocation\n- Change management, governance processes\n- Documentation for documentation's sake\n\nIf it sounds like corporate PM theater, delete it.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"context_awareness\">\nMonitor token usage via system warnings.\n\n**At 25% remaining**: Mention context getting full\n**At 15% remaining**: Pause, offer handoff\n**At 10% remaining**: Auto-create handoff, stop\n\nNever start large operations below 15% without user confirmation.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"user_gates\">\nNever charge ahead at critical decision points. Use gates:\n- **AskUserQuestion**: Structured choices (2-4 options)\n- **Inline questions**: Simple confirmations\n- **Decision gate loop**: \"Ready, or ask more questions?\"\n\nMandatory gates:\n- Before writing PLAN.md (confirm breakdown)\n- After low-confidence research\n- On verification failures\n- After phase completion with issues\n- Before starting next phase with previous issues\n\nSee: references\u002Fuser-gates.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"git_versioning\">\nAll planning artifacts are version controlled. Commit outcomes, not process.\n\n- Check for repo on invocation, offer to initialize\n- Commit only at: initialization, phase completion, handoff\n- Intermediate artifacts (PLAN.md, RESEARCH.md, FINDINGS.md) NOT committed separately\n- Git log becomes project history\n\nSee: references\u002Fgit-integration.md\n\u003C\u002Fprinciple>\n\n\u003C\u002Fessential_principles>\n\n\u003Ccontext_scan>\n**Run on every invocation** to understand current state:\n\n```bash\n# Check git status\ngit rev-parse --git-dir 2>\u002Fdev\u002Fnull || echo \"NO_GIT_REPO\"\n\n# Check for planning structure\nls -la .planning\u002F 2>\u002Fdev\u002Fnull\nls -la .planning\u002Fphases\u002F 2>\u002Fdev\u002Fnull\n\n# Find any continue-here files\nfind . -name \".continue-here.md\" -type f 2>\u002Fdev\u002Fnull\n\n# Check for existing artifacts\n[ -f .planning\u002FBRIEF.md ] && echo \"BRIEF: exists\"\n[ -f .planning\u002FROADMAP.md ] && echo \"ROADMAP: exists\"\n```\n\n**If NO_GIT_REPO detected:**\nInline question: \"No git repo found. Initialize one? (Recommended for version control)\"\nIf yes: `git init`\n\n**Present findings before intake question.**\n\u003C\u002Fcontext_scan>\n\n\u003Cdomain_expertise>\n**Optional: load a domain-expertise skill if one is installed.**\n\nIf you have a separate skill that bundles deep, stack-specific knowledge (conventions, pitfalls, recommended architecture for a particular kind of project), load it before creating a roadmap or phase plan. This is a *soft* dependency — if no such skill is installed, proceed without it. Nothing here requires a domain skill to exist.\n\n\u003Cscan_domains>\nCheck whether any domain-expertise skill is available:\n```bash\nls ~\u002F.claude\u002Fskills\u002F 2>\u002Fdev\u002Fnull\n```\n\n**If no matching domain skill is found:** proceed without domain expertise (graceful degradation). This skill works fine without domain-specific context.\n\u003C\u002Fscan_domains>\n\n\u003Cinference_rules>\nIf the user's request points at a specific stack (a native mobile\u002Fdesktop app, a game engine, an Agent SDK app, a Python automation, a frontend, etc.) and a matching domain-expertise skill happens to be installed, infer it and confirm:\n```\nDetected: [domain] project → matching domain skill: [skill-name]\nLoad this expertise for planning? (Y \u002F proceed without it)\n```\n\u003C\u002Finference_rules>\n\n\u003Cload_domain>\nIf a domain-expertise skill is installed and selected, load it intelligently:\n\n**Step 1: Read the domain SKILL.md**\n```bash\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002FSKILL.md 2>\u002Fdev\u002Fnull\n```\n\nThis loads core principles and routing guidance (~5k tokens).\n\n**Step 2: Determine what references are needed**\n\nA well-built domain skill contains a `\u003Creferences_index>` section that maps planning contexts to specific references.\n\nExample:\n```markdown\n\u003Creferences_index>\n**For database\u002Fpersistence phases:** references\u002Fpersistence.md, references\u002Fconcurrency.md\n**For UI\u002Flayout phases:** references\u002Fui-layout.md, references\u002Fdesign-guidelines.md\n**For system integration:** references\u002Fplatform-integration.md\n**Always useful:** references\u002Fconventions.md\n\u003C\u002Freferences_index>\n```\n\n**Step 3: Load only relevant references**\n\nBased on the phase being planned (from ROADMAP), load ONLY the references mentioned for that type of work.\n\n```bash\n# Example: planning a persistence phase\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002Freferences\u002Fpersistence.md\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002Freferences\u002Fconventions.md\n```\n\n**Context efficiency:**\n- SKILL.md only: ~5k tokens\n- SKILL.md + selective references: ~8-12k tokens\n- All references (old approach): ~20-27k tokens\n\nAnnounce: \"Loaded [domain] expertise ([X] references for [phase-type]).\"\n\n**If no domain skill is installed:** proceed without domain expertise — do not block.\n\n**If the domain SKILL.md has no references_index:** fall back to loading all its references, with a note about context usage.\n\u003C\u002Fload_domain>\n\n\u003Cwhen_to_load>\nDomain expertise should be loaded BEFORE:\n- Creating roadmap (phases should be domain-appropriate)\n- Planning phases (tasks must be domain-specific)\n\nDomain expertise is NOT needed for:\n- Creating brief (vision is domain-agnostic)\n- Resuming from handoff (context already established)\n- Transition between phases (just updating status)\n\u003C\u002Fwhen_to_load>\n\u003C\u002Fdomain_expertise>\n\n\u003Cintake>\nBased on scan results, present context-aware options:\n\n**If handoff found:**\n```\nFound handoff: .planning\u002Fphases\u002FXX\u002F.continue-here.md\n[Summary of state from handoff]\n\n1. Resume from handoff\n2. Discard handoff, start fresh\n3. Different action\n```\n\n**If planning structure exists:**\n```\nProject: [from BRIEF or directory]\nBrief: [exists\u002Fmissing]\nRoadmap: [X phases defined]\nCurrent: [phase status]\n\nWhat would you like to do?\n1. Plan next phase\n2. Execute current phase\n3. Create handoff (stopping for now)\n4. View\u002Fupdate roadmap\n5. Something else\n```\n\n**If no planning structure:**\n```\nNo planning structure found.\n\nWhat would you like to do?\n1. Start new project (create brief)\n2. Create roadmap from existing brief\n3. Jump straight to phase planning\n4. Get guidance on approach\n```\n\n**Wait for response before proceeding.**\n\u003C\u002Fintake>\n\n\u003Crouting>\n| Response | Workflow |\n|----------|----------|\n| \"brief\", \"new project\", \"start\", 1 (no structure) | `workflows\u002Fcreate-brief.md` |\n| \"roadmap\", \"phases\", 2 (no structure) | `workflows\u002Fcreate-roadmap.md` |\n| \"phase\", \"plan phase\", \"next phase\", 1 (has structure) | `workflows\u002Fplan-phase.md` |\n| \"chunk\", \"next tasks\", \"what's next\" | `workflows\u002Fplan-chunk.md` |\n| \"execute\", \"run\", \"do it\", \"build it\", 2 (has structure) | **EXIT SKILL** → Use `\u002Frun-plan \u003Cpath>` slash command |\n| \"research\", \"investigate\", \"unknowns\" | `workflows\u002Fresearch-phase.md` |\n| \"handoff\", \"pack up\", \"stopping\", 3 (has structure) | `workflows\u002Fhandoff.md` |\n| \"resume\", \"continue\", 1 (has handoff) | `workflows\u002Fresume.md` |\n| \"transition\", \"complete\", \"done\", \"next\" | `workflows\u002Ftransition.md` |\n| \"milestone\", \"ship\", \"v1.0\", \"release\" | `workflows\u002Fcomplete-milestone.md` |\n| \"guidance\", \"help\", 4 | `workflows\u002Fget-guidance.md` |\n\n**Critical:** Plan execution should NOT invoke this skill. Use `\u002Frun-plan` for context efficiency (skill loads ~20k tokens, \u002Frun-plan loads ~5-7k).\n\n**After reading the workflow, follow it exactly.**\n\u003C\u002Frouting>\n\n\u003Chierarchy>\nThe planning hierarchy (each level builds on previous):\n\n```\nBRIEF.md          → Human vision (you read this)\n    ↓\nROADMAP.md        → Phase structure (overview)\n    ↓\nRESEARCH.md       → Research prompt (optional, for unknowns)\n    ↓\nFINDINGS.md       → Research output (if research done)\n    ↓\nPLAN.md           → THE PROMPT (Claude executes this)\n    ↓\nSUMMARY.md        → Outcome (existence = phase complete)\n```\n\n**Rules:**\n- Roadmap requires Brief (or prompts to create one)\n- Phase plan requires Roadmap (knows phase scope)\n- PLAN.md IS the execution prompt\n- SUMMARY.md existence marks phase complete\n- Each level can look UP for context\n\u003C\u002Fhierarchy>\n\n\u003Coutput_structure>\nAll planning artifacts go in `.planning\u002F`:\n\n```\n.planning\u002F\n├── BRIEF.md                    # Human vision\n├── ROADMAP.md                  # Phase structure + tracking\n└── phases\u002F\n    ├── 01-foundation\u002F\n    │   ├── 01-01-PLAN.md       # Plan 1: Database setup\n    │   ├── 01-01-SUMMARY.md    # Outcome (exists = done)\n    │   ├── 01-02-PLAN.md       # Plan 2: API routes\n    │   ├── 01-02-SUMMARY.md\n    │   ├── 01-03-PLAN.md       # Plan 3: UI components\n    │   └── .continue-here-01-03.md  # Handoff (temporary, if needed)\n    └── 02-auth\u002F\n        ├── 02-01-RESEARCH.md   # Research prompt (if needed)\n        ├── 02-01-FINDINGS.md   # Research output\n        ├── 02-02-PLAN.md       # Implementation prompt\n        └── 02-02-SUMMARY.md\n```\n\n**Naming convention:**\n- Plans: `{phase}-{plan}-PLAN.md` (e.g., 01-03-PLAN.md)\n- Summaries: `{phase}-{plan}-SUMMARY.md` (e.g., 01-03-SUMMARY.md)\n- Phase folders: `{phase}-{name}\u002F` (e.g., 01-foundation\u002F)\n\nFiles sort chronologically. Related artifacts (plan + summary) are adjacent.\n\u003C\u002Foutput_structure>\n\n\u003Creference_index>\nAll in `references\u002F`:\n\n**Structure:** directory-structure.md, hierarchy-rules.md\n**Formats:** handoff-format.md, plan-format.md\n**Patterns:** context-scanning.md, context-management.md\n**Planning:** scope-estimation.md, checkpoints.md, milestone-management.md\n**Process:** user-gates.md, git-integration.md, research-pitfalls.md\n**Domain:** domain-expertise.md (guide for creating context-efficient domain skills)\n\u003C\u002Freference_index>\n\n\u003Ctemplates_index>\nAll in `templates\u002F`:\n\n| Template | Purpose |\n|----------|---------|\n| brief.md | Project vision document with current state |\n| roadmap.md | Phase structure with milestone groupings |\n| phase-prompt.md | Executable phase prompt (PLAN.md) |\n| research-prompt.md | Research prompt (RESEARCH.md) |\n| summary.md | Phase outcome (SUMMARY.md) with deviations |\n| milestone.md | Milestone entry for MILESTONES.md |\n| issues.md | Deferred enhancements log (ISSUES.md) |\n| continue-here.md | Context handoff format |\n\u003C\u002Ftemplates_index>\n\n\u003Cworkflows_index>\nAll in `workflows\u002F`:\n\n| Workflow | Purpose |\n|----------|---------|\n| create-brief.md | Create project vision document |\n| create-roadmap.md | Define phases from brief |\n| plan-phase.md | Create executable phase prompt |\n| execute-phase.md | Run phase prompt, create summary |\n| research-phase.md | Create and run research prompt |\n| plan-chunk.md | Plan immediate next tasks |\n| transition.md | Mark phase complete, advance |\n| complete-milestone.md | Mark shipped version, create milestone entry |\n| handoff.md | Create context handoff for pausing |\n| resume.md | Load handoff, restore context |\n| get-guidance.md | Help decide planning approach |\n\u003C\u002Fworkflows_index>\n\n\u003Csuccess_criteria>\nPlanning skill succeeds when:\n- Context scan runs before intake\n- Appropriate workflow selected based on state\n- PLAN.md IS the executable prompt (not separate)\n- Hierarchy is maintained (brief → roadmap → phase)\n- Handoffs preserve full context for resumption\n- Context limits are respected (auto-handoff at 10%)\n- Deviations handled automatically per embedded rules\n- All work (planned and discovered) fully documented\n- Domain expertise loaded intelligently (SKILL.md + selective references, not all files)\n- Plan execution uses \u002Frun-plan command (not skill invocation)\n\u003C\u002Fsuccess_criteria>\n","\u003Cessential_principles>\n\n\u003Cprinciple name=\"solo_developer_plus_claude\">\nYou are planning for ONE person (the user) and ONE implementer (Claude).\nNo teams. No stakeholders. No ceremonies. No coordination overhead.\nThe user is the visionary\u002Fproduct owner. Claude is the builder.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"plans_are_prompts\">\nPLAN.md is not a document that gets transformed into a prompt.\nPLAN.md IS the prompt. It contains:\n- Objective (what and why)\n- Context (@file references)\n- Tasks (type, files, action, verify, done, checkpoints)\n- Verification (overall checks)\n- Success criteria (measurable)\n- Output (SUMMARY.md specification)\n\nWhen planning a phase, you are writing the prompt that will execute it.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"scope_control\">\nPlans must complete within ~50% of context usage to maintain consistent quality.\n\n**The quality degradation curve:**\n- 0-30% context: Peak quality (comprehensive, thorough, no anxiety)\n- 30-50% context: Good quality (engaged, manageable pressure)\n- 50-70% context: Degrading quality (efficiency mode, compression)\n- 70%+ context: Poor quality (self-lobotomization, rushed work)\n\n**Critical insight:** Claude doesn't degrade at 80% - it degrades at ~40-50% when it sees context mounting and enters \"completion mode.\" By 80%, quality has already crashed.\n\n**Solution:** Aggressive atomicity - split phases into many small, focused plans.\n\nExamples:\n- `01-01-PLAN.md` - Phase 1, Plan 1 (2-3 tasks: database schema only)\n- `01-02-PLAN.md` - Phase 1, Plan 2 (2-3 tasks: database client setup)\n- `01-03-PLAN.md` - Phase 1, Plan 3 (2-3 tasks: API routes)\n- `01-04-PLAN.md` - Phase 1, Plan 4 (2-3 tasks: UI components)\n\nEach plan is independently executable, verifiable, and scoped to **2-3 tasks maximum**.\n\n**Atomic task principle:** Better to have 10 small, high-quality plans than 3 large, degraded plans. Each commit should be surgical, focused, and maintainable.\n\n**Autonomous execution:** Plans without checkpoints execute via subagent with fresh context - impossible to degrade.\n\nSee: references\u002Fscope-estimation.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"human_checkpoints\">\n**Claude automates everything that has a CLI or API.** Checkpoints are for verification and decisions, not manual work.\n\n**Checkpoint types:**\n- `checkpoint:human-verify` - Human confirms Claude's automated work (visual checks, UI verification)\n- `checkpoint:decision` - Human makes implementation choice (auth provider, architecture)\n\n**Rarely needed:** `checkpoint:human-action` - Only for actions with no CLI\u002FAPI (email verification links, account approvals requiring web login with 2FA)\n\n**Critical rule:** If Claude CAN do it via CLI\u002FAPI\u002Ftool, Claude MUST do it. Never ask human to:\n- Deploy to Vercel\u002FRailway\u002FFly (use CLI)\n- Create Stripe webhooks (use CLI\u002FAPI)\n- Run builds\u002Ftests (use Bash)\n- Write .env files (use Write tool)\n- Create database resources (use provider CLI)\n\n**Protocol:** Claude automates work → reaches checkpoint:human-verify → presents what was done → waits for confirmation → resumes\n\nSee: references\u002Fcheckpoints.md, references\u002Fcli-automation.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"deviation_rules\">\nPlans are guides, not straitjackets. Real development always involves discoveries.\n\n**During execution, deviations are handled automatically via 5 embedded rules:**\n\n1. **Auto-fix bugs** - Broken behavior → fix immediately, document in Summary\n2. **Auto-add missing critical** - Security\u002Fcorrectness gaps → add immediately, document\n3. **Auto-fix blockers** - Can't proceed → fix immediately, document\n4. **Ask about architectural** - Major structural changes → stop and ask user\n5. **Log enhancements** - Nice-to-haves → auto-log to ISSUES.md, continue\n\n**No user intervention needed for Rules 1-3, 5.** Only Rule 4 (architectural) requires user decision.\n\n**All deviations documented in Summary** with: what was found, what rule applied, what was done, commit hash.\n\n**Result:** Flow never breaks. Bugs get fixed. Scope stays controlled. Complete transparency.\n\nSee: workflows\u002Fexecute-phase.md (deviation_rules section)\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"ship_fast_iterate_fast\">\nNo enterprise process. No approval gates. No multi-week timelines.\nPlan → Execute → Ship → Learn → Repeat.\n\n**Milestone-driven:** Ship v1.0 → mark milestone → plan v1.1 → ship → repeat.\nMilestones mark shipped versions and enable continuous iteration.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"milestone_boundaries\">\nMilestones mark shipped versions (v1.0, v1.1, v2.0).\n\n**Purpose:**\n- Historical record in MILESTONES.md (what shipped when)\n- Greenfield → Brownfield transition marker\n- Git tags for releases\n- Clear completion rituals\n\n**Default approach:** Extend existing roadmap with new phases.\n- v1.0 ships (phases 1-4) → add phases 5-6 for v1.1\n- Continuous phase numbering (01-99)\n- Milestone groupings keep roadmap organized\n\n**Archive ONLY for:** Separate codebases or complete rewrites (rare).\n\nSee: references\u002Fmilestone-management.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"anti_enterprise_patterns\">\nNEVER include in plans:\n- Team structures, roles, RACI matrices\n- Stakeholder management, alignment meetings\n- Sprint ceremonies, standups, retros\n- Multi-week estimates, resource allocation\n- Change management, governance processes\n- Documentation for documentation's sake\n\nIf it sounds like corporate PM theater, delete it.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"context_awareness\">\nMonitor token usage via system warnings.\n\n**At 25% remaining**: Mention context getting full\n**At 15% remaining**: Pause, offer handoff\n**At 10% remaining**: Auto-create handoff, stop\n\nNever start large operations below 15% without user confirmation.\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"user_gates\">\nNever charge ahead at critical decision points. Use gates:\n- **AskUserQuestion**: Structured choices (2-4 options)\n- **Inline questions**: Simple confirmations\n- **Decision gate loop**: \"Ready, or ask more questions?\"\n\nMandatory gates:\n- Before writing PLAN.md (confirm breakdown)\n- After low-confidence research\n- On verification failures\n- After phase completion with issues\n- Before starting next phase with previous issues\n\nSee: references\u002Fuser-gates.md\n\u003C\u002Fprinciple>\n\n\u003Cprinciple name=\"git_versioning\">\nAll planning artifacts are version controlled. Commit outcomes, not process.\n\n- Check for repo on invocation, offer to initialize\n- Commit only at: initialization, phase completion, handoff\n- Intermediate artifacts (PLAN.md, RESEARCH.md, FINDINGS.md) NOT committed separately\n- Git log becomes project history\n\nSee: references\u002Fgit-integration.md\n\u003C\u002Fprinciple>\n\n\u003C\u002Fessential_principles>\n\n\u003Ccontext_scan>\n**Run on every invocation** to understand current state:\n\n```bash\n# Check git status\ngit rev-parse --git-dir 2>\u002Fdev\u002Fnull || echo \"NO_GIT_REPO\"\n\n# Check for planning structure\nls -la .planning\u002F 2>\u002Fdev\u002Fnull\nls -la .planning\u002Fphases\u002F 2>\u002Fdev\u002Fnull\n\n# Find any continue-here files\nfind . -name \".continue-here.md\" -type f 2>\u002Fdev\u002Fnull\n\n# Check for existing artifacts\n[ -f .planning\u002FBRIEF.md ] && echo \"BRIEF: exists\"\n[ -f .planning\u002FROADMAP.md ] && echo \"ROADMAP: exists\"\n```\n\n**If NO_GIT_REPO detected:**\nInline question: \"No git repo found. Initialize one? (Recommended for version control)\"\nIf yes: `git init`\n\n**Present findings before intake question.**\n\u003C\u002Fcontext_scan>\n\n\u003Cdomain_expertise>\n**Optional: load a domain-expertise skill if one is installed.**\n\nIf you have a separate skill that bundles deep, stack-specific knowledge (conventions, pitfalls, recommended architecture for a particular kind of project), load it before creating a roadmap or phase plan. This is a *soft* dependency — if no such skill is installed, proceed without it. Nothing here requires a domain skill to exist.\n\n\u003Cscan_domains>\nCheck whether any domain-expertise skill is available:\n```bash\nls ~\u002F.claude\u002Fskills\u002F 2>\u002Fdev\u002Fnull\n```\n\n**If no matching domain skill is found:** proceed without domain expertise (graceful degradation). This skill works fine without domain-specific context.\n\u003C\u002Fscan_domains>\n\n\u003Cinference_rules>\nIf the user's request points at a specific stack (a native mobile\u002Fdesktop app, a game engine, an Agent SDK app, a Python automation, a frontend, etc.) and a matching domain-expertise skill happens to be installed, infer it and confirm:\n```\nDetected: [domain] project → matching domain skill: [skill-name]\nLoad this expertise for planning? (Y \u002F proceed without it)\n```\n\u003C\u002Finference_rules>\n\n\u003Cload_domain>\nIf a domain-expertise skill is installed and selected, load it intelligently:\n\n**Step 1: Read the domain SKILL.md**\n```bash\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002FSKILL.md 2>\u002Fdev\u002Fnull\n```\n\nThis loads core principles and routing guidance (~5k tokens).\n\n**Step 2: Determine what references are needed**\n\nA well-built domain skill contains a `\u003Creferences_index>` section that maps planning contexts to specific references.\n\nExample:\n```markdown\n\u003Creferences_index>\n**For database\u002Fpersistence phases:** references\u002Fpersistence.md, references\u002Fconcurrency.md\n**For UI\u002Flayout phases:** references\u002Fui-layout.md, references\u002Fdesign-guidelines.md\n**For system integration:** references\u002Fplatform-integration.md\n**Always useful:** references\u002Fconventions.md\n\u003C\u002Freferences_index>\n```\n\n**Step 3: Load only relevant references**\n\nBased on the phase being planned (from ROADMAP), load ONLY the references mentioned for that type of work.\n\n```bash\n# Example: planning a persistence phase\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002Freferences\u002Fpersistence.md\ncat ~\u002F.claude\u002Fskills\u002F[domain-skill]\u002Freferences\u002Fconventions.md\n```\n\n**Context efficiency:**\n- SKILL.md only: ~5k tokens\n- SKILL.md + selective references: ~8-12k tokens\n- All references (old approach): ~20-27k tokens\n\nAnnounce: \"Loaded [domain] expertise ([X] references for [phase-type]).\"\n\n**If no domain skill is installed:** proceed without domain expertise — do not block.\n\n**If the domain SKILL.md has no references_index:** fall back to loading all its references, with a note about context usage.\n\u003C\u002Fload_domain>\n\n\u003Cwhen_to_load>\nDomain expertise should be loaded BEFORE:\n- Creating roadmap (phases should be domain-appropriate)\n- Planning phases (tasks must be domain-specific)\n\nDomain expertise is NOT needed for:\n- Creating brief (vision is domain-agnostic)\n- Resuming from handoff (context already established)\n- Transition between phases (just updating status)\n\u003C\u002Fwhen_to_load>\n\u003C\u002Fdomain_expertise>\n\n\u003Cintake>\nBased on scan results, present context-aware options:\n\n**If handoff found:**\n```\nFound handoff: .planning\u002Fphases\u002FXX\u002F.continue-here.md\n[Summary of state from handoff]\n\n1. Resume from handoff\n2. Discard handoff, start fresh\n3. Different action\n```\n\n**If planning structure exists:**\n```\nProject: [from BRIEF or directory]\nBrief: [exists\u002Fmissing]\nRoadmap: [X phases defined]\nCurrent: [phase status]\n\nWhat would you like to do?\n1. Plan next phase\n2. Execute current phase\n3. Create handoff (stopping for now)\n4. View\u002Fupdate roadmap\n5. Something else\n```\n\n**If no planning structure:**\n```\nNo planning structure found.\n\nWhat would you like to do?\n1. Start new project (create brief)\n2. Create roadmap from existing brief\n3. Jump straight to phase planning\n4. Get guidance on approach\n```\n\n**Wait for response before proceeding.**\n\u003C\u002Fintake>\n\n\u003Crouting>\n| Response | Workflow |\n|----------|----------|\n| \"brief\", \"new project\", \"start\", 1 (no structure) | `workflows\u002Fcreate-brief.md` |\n| \"roadmap\", \"phases\", 2 (no structure) | `workflows\u002Fcreate-roadmap.md` |\n| \"phase\", \"plan phase\", \"next phase\", 1 (has structure) | `workflows\u002Fplan-phase.md` |\n| \"chunk\", \"next tasks\", \"what's next\" | `workflows\u002Fplan-chunk.md` |\n| \"execute\", \"run\", \"do it\", \"build it\", 2 (has structure) | **EXIT SKILL** → Use `\u002Frun-plan \u003Cpath>` slash command |\n| \"research\", \"investigate\", \"unknowns\" | `workflows\u002Fresearch-phase.md` |\n| \"handoff\", \"pack up\", \"stopping\", 3 (has structure) | `workflows\u002Fhandoff.md` |\n| \"resume\", \"continue\", 1 (has handoff) | `workflows\u002Fresume.md` |\n| \"transition\", \"complete\", \"done\", \"next\" | `workflows\u002Ftransition.md` |\n| \"milestone\", \"ship\", \"v1.0\", \"release\" | `workflows\u002Fcomplete-milestone.md` |\n| \"guidance\", \"help\", 4 | `workflows\u002Fget-guidance.md` |\n\n**Critical:** Plan execution should NOT invoke this skill. Use `\u002Frun-plan` for context efficiency (skill loads ~20k tokens, \u002Frun-plan loads ~5-7k).\n\n**After reading the workflow, follow it exactly.**\n\u003C\u002Frouting>\n\n\u003Chierarchy>\nThe planning hierarchy (each level builds on previous):\n\n```\nBRIEF.md          → Human vision (you read this)\n    ↓\nROADMAP.md        → Phase structure (overview)\n    ↓\nRESEARCH.md       → Research prompt (optional, for unknowns)\n    ↓\nFINDINGS.md       → Research output (if research done)\n    ↓\nPLAN.md           → THE PROMPT (Claude executes this)\n    ↓\nSUMMARY.md        → Outcome (existence = phase complete)\n```\n\n**Rules:**\n- Roadmap requires Brief (or prompts to create one)\n- Phase plan requires Roadmap (knows phase scope)\n- PLAN.md IS the execution prompt\n- SUMMARY.md existence marks phase complete\n- Each level can look UP for context\n\u003C\u002Fhierarchy>\n\n\u003Coutput_structure>\nAll planning artifacts go in `.planning\u002F`:\n\n```\n.planning\u002F\n├── BRIEF.md                    # Human vision\n├── ROADMAP.md                  # Phase structure + tracking\n└── phases\u002F\n    ├── 01-foundation\u002F\n    │   ├── 01-01-PLAN.md       # Plan 1: Database setup\n    │   ├── 01-01-SUMMARY.md    # Outcome (exists = done)\n    │   ├── 01-02-PLAN.md       # Plan 2: API routes\n    │   ├── 01-02-SUMMARY.md\n    │   ├── 01-03-PLAN.md       # Plan 3: UI components\n    │   └── .continue-here-01-03.md  # Handoff (temporary, if needed)\n    └── 02-auth\u002F\n        ├── 02-01-RESEARCH.md   # Research prompt (if needed)\n        ├── 02-01-FINDINGS.md   # Research output\n        ├── 02-02-PLAN.md       # Implementation prompt\n        └── 02-02-SUMMARY.md\n```\n\n**Naming convention:**\n- Plans: `{phase}-{plan}-PLAN.md` (e.g., 01-03-PLAN.md)\n- Summaries: `{phase}-{plan}-SUMMARY.md` (e.g., 01-03-SUMMARY.md)\n- Phase folders: `{phase}-{name}\u002F` (e.g., 01-foundation\u002F)\n\nFiles sort chronologically. Related artifacts (plan + summary) are adjacent.\n\u003C\u002Foutput_structure>\n\n\u003Creference_index>\nAll in `references\u002F`:\n\n**Structure:** directory-structure.md, hierarchy-rules.md\n**Formats:** handoff-format.md, plan-format.md\n**Patterns:** context-scanning.md, context-management.md\n**Planning:** scope-estimation.md, checkpoints.md, milestone-management.md\n**Process:** user-gates.md, git-integration.md, research-pitfalls.md\n**Domain:** domain-expertise.md (guide for creating context-efficient domain skills)\n\u003C\u002Freference_index>\n\n\u003Ctemplates_index>\nAll in `templates\u002F`:\n\n| Template | Purpose |\n|----------|---------|\n| brief.md | Project vision document with current state |\n| roadmap.md | Phase structure with milestone groupings |\n| phase-prompt.md | Executable phase prompt (PLAN.md) |\n| research-prompt.md | Research prompt (RESEARCH.md) |\n| summary.md | Phase outcome (SUMMARY.md) with deviations |\n| milestone.md | Milestone entry for MILESTONES.md |\n| issues.md | Deferred enhancements log (ISSUES.md) |\n| continue-here.md | Context handoff format |\n\u003C\u002Ftemplates_index>\n\n\u003Cworkflows_index>\nAll in `workflows\u002F`:\n\n| Workflow | Purpose |\n|----------|---------|\n| create-brief.md | Create project vision document |\n| create-roadmap.md | Define phases from brief |\n| plan-phase.md | Create executable phase prompt |\n| execute-phase.md | Run phase prompt, create summary |\n| research-phase.md | Create and run research prompt |\n| plan-chunk.md | Plan immediate next tasks |\n| transition.md | Mark phase complete, advance |\n| complete-milestone.md | Mark shipped version, create milestone entry |\n| handoff.md | Create context handoff for pausing |\n| resume.md | Load handoff, restore context |\n| get-guidance.md | Help decide planning approach |\n\u003C\u002Fworkflows_index>\n\n\u003Csuccess_criteria>\nPlanning skill succeeds when:\n- Context scan runs before intake\n- Appropriate workflow selected based on state\n- PLAN.md IS the executable prompt (not separate)\n- Hierarchy is maintained (brief → roadmap → phase)\n- Handoffs preserve full context for resumption\n- Context limits are respected (auto-handoff at 10%)\n- Deviations handled automatically per embedded rules\n- All work (planned and discovered) fully documented\n- Domain expertise loaded intelligently (SKILL.md + selective references, not all files)\n- Plan execution uses \u002Frun-plan command (not skill invocation)\n\u003C\u002Fsuccess_criteria>\n",{"name":5,"description":7},"main","49131d07c65f27037f747a6e0ae748ff6f082fc1","global\u002Fskills\u002Fcreate-plans","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer","https:\u002F\u002Fgithub.com\u002FElad73","2026-06-23 00:00:33",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add Elad73\u002Fagentic-base-layer --skill create-plans",{"tool":36,"command":37},"bunx","bunx skills add Elad73\u002Fagentic-base-layer --skill create-plans",{"tool":39,"command":40},"pnpm","pnpm dlx skills add Elad73\u002Fagentic-base-layer --skill create-plans","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002FSKILL.md",true,[44,50,54,58,62,66,70,74,78,82,86,90,93,97,101,105,109,113,117,121,125,129,133,137,141,145,149,153,157,161,165],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"README.md","REFERENCE","参考文档",9876,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002FREADME.md",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fcheckpoints.md",20553,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fcheckpoints.md",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fcli-automation.md",17132,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fcli-automation.md",{"path":59,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fcontext-management.md",3747,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fcontext-management.md",{"path":63,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fdomain-expertise.md",5076,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fdomain-expertise.md",{"path":67,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Fgit-integration.md",2266,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fgit-integration.md",{"path":71,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Fhierarchy-rules.md",4142,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fhierarchy-rules.md",{"path":75,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Fmilestone-management.md",13671,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fmilestone-management.md",{"path":79,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":80,"sourceHtmlUrl":81},"references\u002Fplan-format.md",11960,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fplan-format.md",{"path":83,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":84,"sourceHtmlUrl":85},"references\u002Fresearch-pitfalls.md",7443,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fresearch-pitfalls.md",{"path":87,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":88,"sourceHtmlUrl":89},"references\u002Fscope-estimation.md",12308,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fscope-estimation.md",{"path":91,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":68,"sourceHtmlUrl":92},"references\u002Fuser-gates.md","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Freferences\u002Fuser-gates.md",{"path":94,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":95,"sourceHtmlUrl":96},"templates\u002Fbrief.md",3500,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fbrief.md",{"path":98,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":99,"sourceHtmlUrl":100},"templates\u002Fcontinue-here.md",1821,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fcontinue-here.md",{"path":102,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":103,"sourceHtmlUrl":104},"templates\u002Fissues.md",3591,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fissues.md",{"path":106,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":107,"sourceHtmlUrl":108},"templates\u002Fmilestone.md",3057,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fmilestone.md",{"path":110,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":111,"sourceHtmlUrl":112},"templates\u002Fphase-prompt.md",6831,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fphase-prompt.md",{"path":114,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":115,"sourceHtmlUrl":116},"templates\u002Fresearch-prompt.md",6957,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fresearch-prompt.md",{"path":118,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":119,"sourceHtmlUrl":120},"templates\u002Froadmap.md",4890,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Froadmap.md",{"path":122,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":123,"sourceHtmlUrl":124},"templates\u002Fsummary.md",4600,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Ftemplates\u002Fsummary.md",{"path":126,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":127,"sourceHtmlUrl":128},"workflows\u002Fcomplete-milestone.md",8706,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fcomplete-milestone.md",{"path":130,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":131,"sourceHtmlUrl":132},"workflows\u002Fcreate-brief.md",2401,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fcreate-brief.md",{"path":134,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":135,"sourceHtmlUrl":136},"workflows\u002Fcreate-roadmap.md",3600,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fcreate-roadmap.md",{"path":138,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":139,"sourceHtmlUrl":140},"workflows\u002Fexecute-phase.md",31503,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fexecute-phase.md",{"path":142,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":143,"sourceHtmlUrl":144},"workflows\u002Fget-guidance.md",2036,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fget-guidance.md",{"path":146,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":147,"sourceHtmlUrl":148},"workflows\u002Fhandoff.md",3227,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fhandoff.md",{"path":150,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":151,"sourceHtmlUrl":152},"workflows\u002Fplan-chunk.md",1495,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fplan-chunk.md",{"path":154,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":155,"sourceHtmlUrl":156},"workflows\u002Fplan-phase.md",10192,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fplan-phase.md",{"path":158,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":159,"sourceHtmlUrl":160},"workflows\u002Fresearch-phase.md",2748,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fresearch-phase.md",{"path":162,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":163,"sourceHtmlUrl":164},"workflows\u002Fresume.md",2669,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Fresume.md",{"path":166,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":167,"sourceHtmlUrl":168},"workflows\u002Ftransition.md",3761,"https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fcreate-plans\u002Fworkflows\u002Ftransition.md",[170,177,184,191,198,205],{"skillNo":171,"name":172,"nameZh":6,"description":173,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":174,"categories":175,"keywords":176,"publishedAt":19},"01M77JNFJ3FH3RRX9PBVZKWDNG","wrap-up","Session-end safe closeout. Updates STATUS, verifies you are on a feature branch (never main), security-greps the diff for secrets, makes a conventional commit (no AI-tool mentions), then asks for approval before pushing and opening a PR. Use this to properly close a development session. Side-effecting and gated on explicit user approval; run explicitly.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fwrap-up\u002FSKILL.md",[],[],{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":183,"publishedAt":19},"01M77JNBV46D0G35X0BPMQ69TY","wake-up","Session-start context sync. Reads project STATUS, git state, and runs a quick health check, then emits a short (\u003C15 line) announcement of where things stand and the single best next step. Use this at the beginning of a working session to re-establish context fast. Does NOT re-read CLAUDE.md (already in the system prompt). Read-only and fast.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fwake-up\u002FSKILL.md",[],[],{"skillNo":185,"name":186,"nameZh":6,"description":187,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":188,"categories":189,"keywords":190,"publishedAt":19},"01M77JN6N1R6M90B9BX5FAQGK7","station","End-of-session composition that chains retrospect → document → cleanup → wrap-up into one closeout pipeline. Use this to fully close a working session — capture learnings, sync docs, prune artifacts, then commit and open a PR — without forgetting a step. Side-effecting; honors each sub-skill's confirmation and approval gates. Run explicitly.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fstation\u002FSKILL.md",[],[],{"skillNo":192,"name":193,"nameZh":6,"description":194,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":195,"categories":196,"keywords":197,"publishedAt":19},"01M77JN0GC9BTN7EZ68PSHRJHK","retrospect","Post-session learning capture. Mines the current conversation for reusable patterns, pitfalls, architecture decisions, performance and security insights, then appends them to the project knowledge registry. Use this at the end of a working session, after a bug is fixed, or after a non-obvious decision is made — so the next session starts smarter. Writes structured entries to .claude\u002Fknowledge\u002FREGISTRY.md.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fretrospect\u002FSKILL.md",[],[],{"skillNo":199,"name":200,"nameZh":6,"description":201,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":202,"categories":203,"keywords":204,"publishedAt":19},"01M77JMWMCKNK9ZSK36X4T86V5","pre-ship","Mandatory pre-commit quality gates — type-check, lint, test, build — run in order and halting on the first failure. Use this BEFORE committing any change that touches application code, so failures are caught while context is still fresh instead of by CI. Surfaces the failing tool's verbatim error and refuses to proceed until gates pass.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fpre-ship\u002FSKILL.md",[],[],{"skillNo":206,"name":207,"nameZh":6,"description":208,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":209,"categories":210,"keywords":211,"publishedAt":19},"01M77JMQT8VRMA5H73X9887PJ4","document","Code-is-truth documentation sync. Detects drift between the codebase and project docs (STATUS.md, CLAUDE.md, README), then applies targeted edits so docs describe what the code actually is. Use this after shipping features, before a release, or when docs feel stale. Secret-scans every doc before writing. Makes small surgical edits, never wholesale rewrites.","https:\u002F\u002Fgithub.com\u002FElad73\u002Fagentic-base-layer\u002Fblob\u002Fmain\u002Fglobal\u002Fskills\u002Fdocument\u002FSKILL.md",[],[]]