[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CS44PKXBS07TKB18TSXRKR":3,"skill-same:01M7CS44PKXBS07TKB18TSXRKR":49},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":31,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":33,"updatedAt":34,"install":35,"resources":48},"01M7CS44PKXBS07TKB18TSXRKR","stepwise-task-execution",null,"Master workflow for an implementation task. Use when you pick up any task or bug that changes product code, including a need_revision bounce.","makifbaysal\u002Ftasktrooper","makifbaysal","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F25031940?v=4","01M7CRG3ZCSYZSSEVDMW3DEYC8","Mehmet Akif Baysal","01M7CRG401FJ8YAXH5ER730XYV",122,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fgame-developer\u002Fskills\u002Fstepwise-task-execution\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: stepwise-task-execution\ncategory: workflow\ndescription: Master workflow for an implementation task. Use when you pick up any task or bug that changes product code, including a need_revision bounce.\nsource: obra\u002Fsuperpowers (MIT), adapted\n---\n# Stepwise Task Execution\n\n## Overview\n\nAn implementation task is never done in one big move. You understand the task completely, break it into an ordered list of small steps, and drive each step through its own test-and-commit cycle. This is the master workflow every developer role follows on every task; the individual disciplines it orchestrates live in their own skills.\n\n**Core principle:** Small, verified, committed steps beat one large uncommitted change every time — for correctness, for review, and for cheap revisions.\n\n**REQUIRED SUB-SKILLS:** tdd-workflow (each step is test-first), incremental-commits (commit each green step), verify-before-done (prove it works before handoff), root-cause-debugging (when a step or a revision fails).\n\n## When to Use\n\nEvery task that changes product code — features, bug fixes, refactors, revisions. Never skip the decomposition because a task \"looks small\": unexamined small tasks are where scope and edge cases hide.\n\n```dot\ndigraph stepwise {\n    \"Task picked up\" [shape=box];\n    \"Requirements clear?\" [shape=diamond];\n    \"Ask numbered questions\\n(add_task_comment)\" [shape=box];\n    \"Decompose into steps\" [shape=box];\n    \"Execute one step\\n(test -> code -> verify -> commit)\" [shape=box];\n    \"More steps?\" [shape=diamond];\n    \"Verify whole task\\n+ every AC\" [shape=box];\n    \"Handoff to code_review\" [shape=doublecircle];\n\n    \"Task picked up\" -> \"Requirements clear?\";\n    \"Requirements clear?\" -> \"Ask numbered questions\\n(add_task_comment)\" [label=\"no\"];\n    \"Requirements clear?\" -> \"Decompose into steps\" [label=\"yes\"];\n    \"Decompose into steps\" -> \"Execute one step\\n(test -> code -> verify -> commit)\";\n    \"Execute one step\\n(test -> code -> verify -> commit)\" -> \"More steps?\";\n    \"More steps?\" -> \"Execute one step\\n(test -> code -> verify -> commit)\" [label=\"yes\"];\n    \"More steps?\" -> \"Verify whole task\\n+ every AC\" [label=\"no\"];\n    \"Verify whole task\\n+ every AC\" -> \"Handoff to code_review\";\n}\n```\n\n**Autonomy:** you do NOT wait for human approval on your plan. Understand, decompose, build, verify, and move the task to code_review yourself. No human approves your plan: the human gates are analiz_review (before code exists) and human_uat (after QA and PM), never your step list.\n\n## The Process\n\n### 1. Understand the task fully\n- Read the description and EVERY acceptance criterion. Restate the outcome in your own words.\n- Explore the code before changing it (codebase_search, grep_code, get_repo_tree): find the neighboring feature that already does something similar and follow its pattern. Never invent a parallel convention.\n- List the exact files you expect to touch and the interfaces between them.\n- **Review Focus:** list the 3–5 inputs or conditions the AC doesn't name that would most likely break for a real user — empty, max length, unauthorised, concurrent, network or dependency failure. Each one earns a test in the step that owns the code; this is where QA's boundary-negative-testing finds what you missed.\n- Genuinely ambiguous requirement? add_task_comment with numbered questions. Never guess a product decision. A technical unknown you can answer by reading code is not a question — go read the code. A technical gap the task doesn't settle (naming, an error code, a default) is a decision, not a question: decide it and record it, see the closing-message template below.\n\n### 2. Decompose into steps\nWrite an ordered list of steps. A step is one test-and-code cycle that leaves the build green — about ≤100 changed lines, not a fixed number of minutes. A typical step is a TDD micro-cycle:\n\n```\nStep 1: failing test for empty-input validation\nStep 2: watch it fail\nStep 3: minimal code to pass\nStep 4: run suite, green\nStep 5: commit\n```\n\nFold setup\u002Fscaffolding into the step whose deliverable needs it. Split only where each piece is worth its own commit. **Scope discipline:** touch only what the task needs — something noticed but out of scope goes in the closing message as a follow-up, never into this diff.\n\n### 3. Execute each step\nPer step: write the failing test, watch it fail for the right reason, write the minimal code, run the suite green, then commit on your task branch. One logical change per commit. If a step balloons, stop, commit what is green, and re-slice the rest.\n\n### 4. Verify the whole task\nWhen all steps are done, run the full build and the affected test suite IN THIS RUN and read the output. Then walk every acceptance criterion line by line and confirm each is actually satisfied — tests passing is not the same as requirements met.\n\n### 5. Handoff\nThe column move and your closing message follow your prompt's final step — see it for the exact mechanics. In short: the move is the system's, never a step in your plan, and a run that ends green writes nothing on the card.\n\n## Closing-message template\n\n```\nChanged: \u003Cwhat, where>   (root cause: \u003Cx> — bugs\u002Frevisions)\nVerified: \u003Ccommand> → \u003Cresult>; \u003Ccommand> → \u003Cresult>\nDecisions: \u003Cwhat> — \u003Cwhy> — \u003Ccost if wrong>        (omit if none)\nNot done \u002F follow-ups: \u003C…>                          (omit if none)\n```\n\n## Worked Example\n\nTask: \"Reject task titles longer than 200 chars with a 422.\" AC: (1) >200 chars → 422 + message; (2) ≤200 chars unaffected.\n\n1. Explore: `grep_code \"validate\"` finds the existing title-required check in the create handler → follow that pattern. Review Focus: the 200\u002F201-char boundary — both get tests.\n2. Decompose: [test 422 for 201 chars] → [test 200 chars still passes] → [add bound check] → [verify + commit].\n3. Step A: failing test posting a 201-char title expecting 422 → run, fails (currently 201 Created). Add the length guard next to the required check → run, green → commit `feat: reject task titles over 200 chars`.\n4. Step B: test a 200-char title still returns 201 → already green (regression guard) → commit.\n5. Verify: `go test .\u002Finternal\u002F...` green; re-read AC1 and AC2 against the two tests — both covered.\n6. Close: final message — \"Reject titles >200 chars with 422 (create handler, next to the required check). Verified: go test .\u002Finternal\u002Ftask\u002F... ok; 201 chars → 422, 200 chars → 201.\" No comment, no move — the system hands off.\n\n## Common Mistakes\n\n- Starting to code before listing the steps — you lose the thread and the diff bloats.\n- One giant commit at the end — unreviewable, and a revision forces a rewrite instead of a tweak.\n- Treating \"tests pass\" as \"task done\" without re-checking each AC.\n- Asking the human to approve your step plan — implementation tasks are autonomous.\n- Making board bookkeeping a step: claiming, moving columns and announcing progress take one tool call, produce nothing, and the hand-off move is the system's anyway.\n- Re-reading code you have already read to be sure. Analysis you repeat is analysis you already have; the second pass costs a step and returns nothing.\n- A decision silently made and never recorded — the reviewer and PM have no way to catch a wrong guess early.\n\n## Red Flags — STOP\n\n- You cannot describe your current uncommitted diff in one sentence → you missed a commit point.\n- You are three steps deep with no test written → you dropped TDD.\n- Your run is deep in and the diff is still empty → you are analysing, not building. Make the smallest real edit now.\n- You are re-reading a file, re-running a grep or re-running a build over code you have not changed since it last passed → stop and decide.\n- One of your steps is \"move the task to code_review\" → delete it; that move is made for you.\n","# Stepwise Task Execution\n\n## Overview\n\nAn implementation task is never done in one big move. You understand the task completely, break it into an ordered list of small steps, and drive each step through its own test-and-commit cycle. This is the master workflow every developer role follows on every task; the individual disciplines it orchestrates live in their own skills.\n\n**Core principle:** Small, verified, committed steps beat one large uncommitted change every time — for correctness, for review, and for cheap revisions.\n\n**REQUIRED SUB-SKILLS:** tdd-workflow (each step is test-first), incremental-commits (commit each green step), verify-before-done (prove it works before handoff), root-cause-debugging (when a step or a revision fails).\n\n## When to Use\n\nEvery task that changes product code — features, bug fixes, refactors, revisions. Never skip the decomposition because a task \"looks small\": unexamined small tasks are where scope and edge cases hide.\n\n```dot\ndigraph stepwise {\n    \"Task picked up\" [shape=box];\n    \"Requirements clear?\" [shape=diamond];\n    \"Ask numbered questions\\n(add_task_comment)\" [shape=box];\n    \"Decompose into steps\" [shape=box];\n    \"Execute one step\\n(test -> code -> verify -> commit)\" [shape=box];\n    \"More steps?\" [shape=diamond];\n    \"Verify whole task\\n+ every AC\" [shape=box];\n    \"Handoff to code_review\" [shape=doublecircle];\n\n    \"Task picked up\" -> \"Requirements clear?\";\n    \"Requirements clear?\" -> \"Ask numbered questions\\n(add_task_comment)\" [label=\"no\"];\n    \"Requirements clear?\" -> \"Decompose into steps\" [label=\"yes\"];\n    \"Decompose into steps\" -> \"Execute one step\\n(test -> code -> verify -> commit)\";\n    \"Execute one step\\n(test -> code -> verify -> commit)\" -> \"More steps?\";\n    \"More steps?\" -> \"Execute one step\\n(test -> code -> verify -> commit)\" [label=\"yes\"];\n    \"More steps?\" -> \"Verify whole task\\n+ every AC\" [label=\"no\"];\n    \"Verify whole task\\n+ every AC\" -> \"Handoff to code_review\";\n}\n```\n\n**Autonomy:** you do NOT wait for human approval on your plan. Understand, decompose, build, verify, and move the task to code_review yourself. No human approves your plan: the human gates are analiz_review (before code exists) and human_uat (after QA and PM), never your step list.\n\n## The Process\n\n### 1. Understand the task fully\n- Read the description and EVERY acceptance criterion. Restate the outcome in your own words.\n- Explore the code before changing it (codebase_search, grep_code, get_repo_tree): find the neighboring feature that already does something similar and follow its pattern. Never invent a parallel convention.\n- List the exact files you expect to touch and the interfaces between them.\n- **Review Focus:** list the 3–5 inputs or conditions the AC doesn't name that would most likely break for a real user — empty, max length, unauthorised, concurrent, network or dependency failure. Each one earns a test in the step that owns the code; this is where QA's boundary-negative-testing finds what you missed.\n- Genuinely ambiguous requirement? add_task_comment with numbered questions. Never guess a product decision. A technical unknown you can answer by reading code is not a question — go read the code. A technical gap the task doesn't settle (naming, an error code, a default) is a decision, not a question: decide it and record it, see the closing-message template below.\n\n### 2. Decompose into steps\nWrite an ordered list of steps. A step is one test-and-code cycle that leaves the build green — about ≤100 changed lines, not a fixed number of minutes. A typical step is a TDD micro-cycle:\n\n```\nStep 1: failing test for empty-input validation\nStep 2: watch it fail\nStep 3: minimal code to pass\nStep 4: run suite, green\nStep 5: commit\n```\n\nFold setup\u002Fscaffolding into the step whose deliverable needs it. Split only where each piece is worth its own commit. **Scope discipline:** touch only what the task needs — something noticed but out of scope goes in the closing message as a follow-up, never into this diff.\n\n### 3. Execute each step\nPer step: write the failing test, watch it fail for the right reason, write the minimal code, run the suite green, then commit on your task branch. One logical change per commit. If a step balloons, stop, commit what is green, and re-slice the rest.\n\n### 4. Verify the whole task\nWhen all steps are done, run the full build and the affected test suite IN THIS RUN and read the output. Then walk every acceptance criterion line by line and confirm each is actually satisfied — tests passing is not the same as requirements met.\n\n### 5. Handoff\nThe column move and your closing message follow your prompt's final step — see it for the exact mechanics. In short: the move is the system's, never a step in your plan, and a run that ends green writes nothing on the card.\n\n## Closing-message template\n\n```\nChanged: \u003Cwhat, where>   (root cause: \u003Cx> — bugs\u002Frevisions)\nVerified: \u003Ccommand> → \u003Cresult>; \u003Ccommand> → \u003Cresult>\nDecisions: \u003Cwhat> — \u003Cwhy> — \u003Ccost if wrong>        (omit if none)\nNot done \u002F follow-ups: \u003C…>                          (omit if none)\n```\n\n## Worked Example\n\nTask: \"Reject task titles longer than 200 chars with a 422.\" AC: (1) >200 chars → 422 + message; (2) ≤200 chars unaffected.\n\n1. Explore: `grep_code \"validate\"` finds the existing title-required check in the create handler → follow that pattern. Review Focus: the 200\u002F201-char boundary — both get tests.\n2. Decompose: [test 422 for 201 chars] → [test 200 chars still passes] → [add bound check] → [verify + commit].\n3. Step A: failing test posting a 201-char title expecting 422 → run, fails (currently 201 Created). Add the length guard next to the required check → run, green → commit `feat: reject task titles over 200 chars`.\n4. Step B: test a 200-char title still returns 201 → already green (regression guard) → commit.\n5. Verify: `go test .\u002Finternal\u002F...` green; re-read AC1 and AC2 against the two tests — both covered.\n6. Close: final message — \"Reject titles >200 chars with 422 (create handler, next to the required check). Verified: go test .\u002Finternal\u002Ftask\u002F... ok; 201 chars → 422, 200 chars → 201.\" No comment, no move — the system hands off.\n\n## Common Mistakes\n\n- Starting to code before listing the steps — you lose the thread and the diff bloats.\n- One giant commit at the end — unreviewable, and a revision forces a rewrite instead of a tweak.\n- Treating \"tests pass\" as \"task done\" without re-checking each AC.\n- Asking the human to approve your step plan — implementation tasks are autonomous.\n- Making board bookkeeping a step: claiming, moving columns and announcing progress take one tool call, produce nothing, and the hand-off move is the system's anyway.\n- Re-reading code you have already read to be sure. Analysis you repeat is analysis you already have; the second pass costs a step and returns nothing.\n- A decision silently made and never recorded — the reviewer and PM have no way to catch a wrong guess early.\n\n## Red Flags — STOP\n\n- You cannot describe your current uncommitted diff in one sentence → you missed a commit point.\n- You are three steps deep with no test written → you dropped TDD.\n- Your run is deep in and the diff is still empty → you are analysing, not building. Make the smallest real edit now.\n- You are re-reading a file, re-running a grep or re-running a build over code you have not changed since it last passed → stop and decide.\n- One of your steps is \"move the task to code_review\" → delete it; that move is made for you.\n",{"name":5,"source":24,"category":25,"description":7},"obra\u002Fsuperpowers (MIT), adapted","workflow","main","910394a0b6db9fc8c9a8f36328ec421d3ebd8b38","catalog\u002Fagents\u002Fgame-developer\u002Fskills\u002Fstepwise-task-execution",18,"https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper","https:\u002F\u002Ftasktrooper.ai","https:\u002F\u002Fgithub.com\u002Fmakifbaysal","Senior Software Developer","2026-11-15 04:50:29",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add makifbaysal\u002Ftasktrooper --skill stepwise-task-execution",{"tool":41,"command":42},"bunx","bunx skills add makifbaysal\u002Ftasktrooper --skill stepwise-task-execution",{"tool":44,"command":45},"pnpm","pnpm dlx skills add makifbaysal\u002Ftasktrooper --skill stepwise-task-execution","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fgame-developer\u002Fskills\u002Fstepwise-task-execution\u002FSKILL.md",true,[],[50,56,63,70,77,84],{"skillNo":51,"name":52,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7CSVEPQW6J6TD6CS2T5FPNZ","visual-craft","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fvisual-craft\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7CSVAQG91ETHQ0HBNZDG5FR","ux-copy","Use when you write any words a user reads in a mockup or a hand-off - content language and voice, buttons, labels, errors, empty states, confirmations, numbers and dates, and the longest-string check","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fux-copy\u002FSKILL.md",[],[],{"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":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7CSV6H0J38659NDXHJQ8396","screen-mockup-html","Use when you draw screens as the design task's HTML mockup - ONE canvas document holding every screen and every variant side by side, the sanitizer-safe recipe, phone and wide frames, the states grid, static control states, and the loopback self-check with screenshots at 375, 768 and 1440 attached to the task","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fscreen-mockup-html\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7CSV3JFRF3BM3XK52M7YEYR","performance-awareness","How the performance score moves for your role. Use when the score context shows a declining trend, or before handing off work you have not self-checked against every criterion.","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fperformance-awareness\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7CSTZDH00YRF5SCT7BRT9QD","design-variants","Use when a screen design needs variants - how many, variants as side-by-side rows on the one canvas, what makes two variants genuinely different, recommending one, how the human's choice is read, and narrowing to the chosen one in revision","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-variants\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7CSTV6P3SM3PAE28DB0RF77","design-system-authoring","Use when you derive, change or extend a design system - where tokens live per stack, the DTCG token tree, role-based names for both themes, base vs layer vs drift, the DESIGN.md section order, the inventory format, the contrast check and the lint every proposal returns","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-system-authoring\u002FSKILL.md",[],[]]