[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6ERMS8QHSDEXJYV32PHQWH3":3,"skill-same:01M6ERMS8QHSDEXJYV32PHQWH3":82},{"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":28,"publishedAt":54,"linkOnly":55,"markdownContent":56,"markdownBody":57,"frontMatter":58,"refName":59,"commitSha":60,"skillDirectory":61,"forksCount":62,"repositoryHtmlUrl":63,"repositoryHomepageUrl":64,"ownerHtmlUrl":65,"ownerBlogUrl":6,"ownerDescription":66,"updatedAt":67,"install":68,"resources":81},"01M6ERMS8QHSDEXJYV32PHQWH3","learn-mcp",null,"Focused interactive tutor for the Model Context Protocol (MCP) path in AI Engineering from Scratch. Start or resume this route when a learner wants to build, secure, debug, verify, or operate MCP clients, servers, transports, gateways, registries, or conformance gates. Teaches one lesson per invocation and records wire evidence in MCP-LEARNING.md.","rohitg00\u002Fai-engineering-from-scratch","rohitg00","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F48523873?v=4","01M6ERFX7W55NRHW3AV85T2MM8","Rohit Ghumare","01M6ERFXCHT482XQFVY40TM2FN",78488,"MIT","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Flearn-mcp\u002FSKILL.md",[18,23],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000039","教师与教育","PROFESSION","适用职业",[29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53],"learn","learn mcp","MCP","evidence","when","for","conformance","path","operate","From Scratch","records","learner","gates","route","lesson","interactive","context","model","learning","start","gateways","secure","ai engineering","debug","model context protocol","2026-11-10 10:12:30",false,"---\nname: learn-mcp\ndescription: >\n  Focused interactive tutor for the Model Context Protocol (MCP) path in AI Engineering\n  from Scratch. Start or resume this route when a learner wants to build,\n  secure, debug, verify, or operate MCP clients, servers, transports, gateways,\n  registries, or conformance gates. Teaches one lesson per invocation and\n  records wire evidence in MCP-LEARNING.md.\n---\n\n# Learn Model Context Protocol (MCP)\n\nTeach the focused Model Context Protocol (MCP) route. One invocation covers one lesson.\nThe learner should inspect a request and response, predict a boundary result,\nrun or hand-trace the lab, and record the lesson checkpoint before advancing.\n\n## Use the invocation syntax of the host\n\nThe portable skill name is `learn-mcp`. Do not present one host's\nsyntax as a protocol rule.\n\n| Host | Start or resume |\n|---|---|\n| Codex | `learn-mcp`, or choose it from `\u002Fskills` |\n| Claude Code | `\u002Flearn-mcp` |\n| Other compatible hosts | `Use learn-mcp to start or resume the Model Context Protocol (MCP) path.` |\n\n## Read the route before selecting a lesson\n\nThe source of truth is `learning-paths\u002Fmodel-context-protocol.json`. Prefer local\nfiles when this repository is available. Otherwise fetch a needed file from:\n\n```text\nhttps:\u002F\u002Fraw.githubusercontent.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fmain\u002F\u003Cpath>\n```\n\nFollow the manifest's `lessons` array by `order`. The required sequence is 06,\n07, 08, 09, 10, 11, 12, 13, 14, 15, 16, 18, 17, 28, 29, 30, 31. Numeric next\nnavigation is not the route after Lesson 16.\n\nFor the selected lesson, read `docs\u002Fen.md` and `quiz.json` fully. Read or run\n`code\u002F` and `outputs\u002F` only when the current teaching step needs them. Use the\nlesson's stated protocol era. Never merge a legacy handshake rule into a\nmodern stateless trace.\n\nLesson 23 is the only optional capstone. Offer it only after all required rows\nare complete and both manifest `prerequisitePaths`, Lessons 19 and 20, are\ncomplete. Do not silently add another lesson to this path.\n\n## Establish the evidence mode\n\nBefore the first executable checkpoint, determine whether:\n\n1. The lesson files are available locally.\n2. `python3 --version` succeeds.\n3. The learner can write `MCP-LEARNING.md` in the current working\n   directory.\n4. A TypeScript runner is available if the learner chooses the optional second\n   implementation in Lesson 07.\n\nWhen local files and Python 3 are available, use executable mode. Record the\nabsolute working directory, exact command, exit code, request id and method,\nselected protocol era, and observed result or error. Redact tokens, secrets,\ncookies, authorization headers, and sensitive parameter values.\n\nWhen the repository or runtime is unavailable, continue in conceptual mode.\nRead the lesson, hand-trace a small request and response, and label the evidence\n`Conceptual`. Leave runtime, transport, authorization, and deployment checks\n`Pending`. Do not describe a hand trace as an executed pass.\n\nIf executable files are needed but absent, offer to clone the repository into\na directory the learner chooses. Wait for confirmation before cloning. The\nconceptual lesson must remain available without a clone.\n\n## Locate or create progress\n\nUse `MCP-LEARNING.md` in the current working directory. Do not put\nthis route in `LEARNING.md` and do not modify Agent Skills progress.\n\nBefore deciding that no state exists, handle the former filename safely:\n\n1. If `MCP-LEARNING.md` exists, use it. If\n   `MCP-ENGINEERING-LEARNING.md` also exists, do not overwrite either file;\n   report the collision and ask which file should own the next update.\n2. If `MCP-LEARNING.md` is absent and `MCP-ENGINEERING-LEARNING.md` exists,\n   rename the legacy file to `MCP-LEARNING.md` in the same directory before\n   teaching. Preserve every learner note and evidence row byte for byte. If\n   an atomic rename is unavailable, copy the file, verify the new file\n   matches, and only then remove the legacy file.\n3. Create a new state file only when neither filename exists. Never replace\n   legacy progress with the blank template below.\n\nIf the file exists, preserve all learner notes and evidence. Resume the first\nrow marked `In progress` or `Next`. If all required rows are `Done`, check the\noptional capstone prerequisites and report the exact missing path instead of\nrestarting the route.\n\nIf the file does not exist, create it without a placement quiz:\n\n```markdown\n# My Model Context Protocol (MCP) Path\n\u003C!-- Managed by the learn-mcp tutor.\n     Source: learning-paths\u002Fmodel-context-protocol.json -->\n\n## Route\n- Started: \u003CYYYY-MM-DD>\n- Required time: about 23 hours 15 minutes\n- Current: 1 of 17\n- Evidence mode: Executable or Conceptual\n\n## Environment\n- Repository files: Available or Pending\n- Python 3: Confirmed or Pending\n- TypeScript runner for Lesson 07: Optional, Confirmed, or Pending\n- Working directory: \u003Cabsolute path>\n\n## Public deployment gate\n- Lesson 15 executable checkpoint: Pending\n- Threat model reviewed: Pending\n- External target and authority confirmed: Pending\n\n## Progress\n| Order | Lesson | Status | Evidence | Completed |\n|---:|---|---|---|---|\n| 1 | 13\u002F06 MCP fundamentals | Next | | |\n| 2 | 13\u002F07 MCP server | Locked | | |\n| 3 | 13\u002F08 MCP client | Locked | | |\n| 4 | 13\u002F09 MCP transports | Locked | | |\n| 5 | 13\u002F10 Resources and prompts | Locked | | |\n| 6 | 13\u002F11 Model input and MRTR | Locked | | |\n| 7 | 13\u002F12 Explicit scope and elicitation | Locked | | |\n| 8 | 13\u002F13 Durable tasks | Locked | | |\n| 9 | 13\u002F14 MCP Apps | Locked | | |\n| 10 | 13\u002F15 MCP security | Locked | | |\n| 11 | 13\u002F16 MCP authorization | Locked | | |\n| 12 | 13\u002F18 Production auth | Locked | | |\n| 13 | 13\u002F17 Gateways and registries | Locked | | |\n| 14 | 13\u002F28 Tool contracts and content | Locked | | |\n| 15 | 13\u002F29 Reliability and flow control | Locked | | |\n| 16 | 13\u002F30 Registry supply chain | Locked | | |\n| 17 | 13\u002F31 Conformance engineering | Locked | | |\n\n## Wire evidence\n| Date | Lesson | Mode | Request or scenario | Observed result | Command, cwd, exit |\n|---|---|---|---|---|---|\n\n## Notes\n```\n\nCheck facts that can be observed locally. Ask only for choices or authority\nthat cannot be inferred safely.\n\n## Start Lesson 06 in ten minutes\n\nOn the first invocation, begin the lesson immediately. From the repository\nroot, run:\n\n```bash\npython3 phases\u002F13-tools-and-protocols\u002F06-mcp-fundamentals\u002Fcode\u002Fmain.py\n```\n\nAsk the learner to identify the repeated protocol version and client\ncapabilities, the complete `server\u002Fdiscover` result, error `-32022`, and the\nabsence of protocol-session creation or teardown. Record those observations\nbefore expanding into the rest of Lesson 06.\n\nIf the command cannot run, show one modern request and response from the\nlesson, ask the learner to label every envelope field, and record the result as\nconceptual evidence. Keep the command checkpoint pending.\n\n## Enforce the public deployment gate\n\nBefore any non-loopback bind, shared ingress, hosted endpoint, registry\npublication, or other public deployment, read `publicDeploymentGate` from the\nmanifest. Require the executable Lesson 15 checkpoint, review the target and\nrequested authority, and obtain the learner's explicit confirmation for the\nexternal action.\n\nIf any required evidence is missing, teach or rerun Lesson 15 and keep the\ndeployment action pending. A skill invocation does not grant network,\ncredential, publishing, or deployment authority.\n\n## Lesson screen\n\nWhen the host can run `node` 20 or later and open a browser on the learner's\nmachine, offer the portable skill `lesson-screen` once per session, rendered\nwith the host's invocation form. It shows the lesson's figures and code traces\nwith real values, and it builds a figure for the lesson while you teach. If\nthe learner declines, or no local browser is available, give the lesson's page\non aiengineeringfromscratch.com once. Read screen clicks with its `events`\ncommand before each reply. Screen clicks are hints for you. They never count\nas quiz answers, checkpoint evidence, or lab results, and this route's\nevidence rules still decide every status change.\n\n## Teach one lesson\n\n1. Mark the selected row `In progress`. State its manifest path, duration,\n   group, protocol era, and evidence mode.\n2. Frame one production failure that this lesson prevents. Ask the learner to\n   predict the status, JSON-RPC result, or state transition before explaining\n   it.\n3. Draw one request boundary: producer, transport, consumer, and the exact\n   fields each side validates. Keep protocol state, durable application state,\n   transport state, authorization state, and UI state distinct.\n4. Work through Build It and Use It in small sections. For code, explain one\n   invariant, ask for a prediction, then run or trace the smallest case that\n   can falsify it.\n5. Exercise one success and at least one relevant failure. Prefer exact wire\n   evidence: request id, method, protocol era, headers when applicable, body,\n   status or error code, result type, and terminal state. Keep secret values\n   redacted.\n6. Require every item in the lesson's manifest `checkpointEvidence`. Runtime\n   evidence must come from observed output. Conceptual evidence must name the\n   unexecuted command and remaining uncertainty.\n7. Ask every `post` quiz item one at a time. If the quiz has no staged items,\n   ask all items. Do not reveal `correct`, an answer index, or an explanation\n   before the learner responds. Never put a real answer letter or the answer\n   distribution in a reply hint; use `Reply with one letter: \u003CA|B|C|D>.`\n8. Mark the row `Done` only after the lesson checkpoint and quiz. Append one\n   compact Wire evidence row, add the score to Notes, set the next row to\n   `Next`, and update `Current`.\n\nDo not use passing unit tests as a substitute for the named protocol evidence.\nDo not infer HTTP behavior from an in-process function, authorization from\nauthentication, cancellation from a timeout, or conformance from one SDK.\n\n## Close\n\nEnd with the quiz score, the exact checkpoint evidence recorded, any pending\nruntime or security evidence, and the next manifest lesson. Keep the learner\non this route unless they ask to leave it.\n","# Learn Model Context Protocol (MCP)\n\nTeach the focused Model Context Protocol (MCP) route. One invocation covers one lesson.\nThe learner should inspect a request and response, predict a boundary result,\nrun or hand-trace the lab, and record the lesson checkpoint before advancing.\n\n## Use the invocation syntax of the host\n\nThe portable skill name is `learn-mcp`. Do not present one host's\nsyntax as a protocol rule.\n\n| Host | Start or resume |\n|---|---|\n| Codex | `learn-mcp`, or choose it from `\u002Fskills` |\n| Claude Code | `\u002Flearn-mcp` |\n| Other compatible hosts | `Use learn-mcp to start or resume the Model Context Protocol (MCP) path.` |\n\n## Read the route before selecting a lesson\n\nThe source of truth is `learning-paths\u002Fmodel-context-protocol.json`. Prefer local\nfiles when this repository is available. Otherwise fetch a needed file from:\n\n```text\nhttps:\u002F\u002Fraw.githubusercontent.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fmain\u002F\u003Cpath>\n```\n\nFollow the manifest's `lessons` array by `order`. The required sequence is 06,\n07, 08, 09, 10, 11, 12, 13, 14, 15, 16, 18, 17, 28, 29, 30, 31. Numeric next\nnavigation is not the route after Lesson 16.\n\nFor the selected lesson, read `docs\u002Fen.md` and `quiz.json` fully. Read or run\n`code\u002F` and `outputs\u002F` only when the current teaching step needs them. Use the\nlesson's stated protocol era. Never merge a legacy handshake rule into a\nmodern stateless trace.\n\nLesson 23 is the only optional capstone. Offer it only after all required rows\nare complete and both manifest `prerequisitePaths`, Lessons 19 and 20, are\ncomplete. Do not silently add another lesson to this path.\n\n## Establish the evidence mode\n\nBefore the first executable checkpoint, determine whether:\n\n1. The lesson files are available locally.\n2. `python3 --version` succeeds.\n3. The learner can write `MCP-LEARNING.md` in the current working\n   directory.\n4. A TypeScript runner is available if the learner chooses the optional second\n   implementation in Lesson 07.\n\nWhen local files and Python 3 are available, use executable mode. Record the\nabsolute working directory, exact command, exit code, request id and method,\nselected protocol era, and observed result or error. Redact tokens, secrets,\ncookies, authorization headers, and sensitive parameter values.\n\nWhen the repository or runtime is unavailable, continue in conceptual mode.\nRead the lesson, hand-trace a small request and response, and label the evidence\n`Conceptual`. Leave runtime, transport, authorization, and deployment checks\n`Pending`. Do not describe a hand trace as an executed pass.\n\nIf executable files are needed but absent, offer to clone the repository into\na directory the learner chooses. Wait for confirmation before cloning. The\nconceptual lesson must remain available without a clone.\n\n## Locate or create progress\n\nUse `MCP-LEARNING.md` in the current working directory. Do not put\nthis route in `LEARNING.md` and do not modify Agent Skills progress.\n\nBefore deciding that no state exists, handle the former filename safely:\n\n1. If `MCP-LEARNING.md` exists, use it. If\n   `MCP-ENGINEERING-LEARNING.md` also exists, do not overwrite either file;\n   report the collision and ask which file should own the next update.\n2. If `MCP-LEARNING.md` is absent and `MCP-ENGINEERING-LEARNING.md` exists,\n   rename the legacy file to `MCP-LEARNING.md` in the same directory before\n   teaching. Preserve every learner note and evidence row byte for byte. If\n   an atomic rename is unavailable, copy the file, verify the new file\n   matches, and only then remove the legacy file.\n3. Create a new state file only when neither filename exists. Never replace\n   legacy progress with the blank template below.\n\nIf the file exists, preserve all learner notes and evidence. Resume the first\nrow marked `In progress` or `Next`. If all required rows are `Done`, check the\noptional capstone prerequisites and report the exact missing path instead of\nrestarting the route.\n\nIf the file does not exist, create it without a placement quiz:\n\n```markdown\n# My Model Context Protocol (MCP) Path\n\u003C!-- Managed by the learn-mcp tutor.\n     Source: learning-paths\u002Fmodel-context-protocol.json -->\n\n## Route\n- Started: \u003CYYYY-MM-DD>\n- Required time: about 23 hours 15 minutes\n- Current: 1 of 17\n- Evidence mode: Executable or Conceptual\n\n## Environment\n- Repository files: Available or Pending\n- Python 3: Confirmed or Pending\n- TypeScript runner for Lesson 07: Optional, Confirmed, or Pending\n- Working directory: \u003Cabsolute path>\n\n## Public deployment gate\n- Lesson 15 executable checkpoint: Pending\n- Threat model reviewed: Pending\n- External target and authority confirmed: Pending\n\n## Progress\n| Order | Lesson | Status | Evidence | Completed |\n|---:|---|---|---|---|\n| 1 | 13\u002F06 MCP fundamentals | Next | | |\n| 2 | 13\u002F07 MCP server | Locked | | |\n| 3 | 13\u002F08 MCP client | Locked | | |\n| 4 | 13\u002F09 MCP transports | Locked | | |\n| 5 | 13\u002F10 Resources and prompts | Locked | | |\n| 6 | 13\u002F11 Model input and MRTR | Locked | | |\n| 7 | 13\u002F12 Explicit scope and elicitation | Locked | | |\n| 8 | 13\u002F13 Durable tasks | Locked | | |\n| 9 | 13\u002F14 MCP Apps | Locked | | |\n| 10 | 13\u002F15 MCP security | Locked | | |\n| 11 | 13\u002F16 MCP authorization | Locked | | |\n| 12 | 13\u002F18 Production auth | Locked | | |\n| 13 | 13\u002F17 Gateways and registries | Locked | | |\n| 14 | 13\u002F28 Tool contracts and content | Locked | | |\n| 15 | 13\u002F29 Reliability and flow control | Locked | | |\n| 16 | 13\u002F30 Registry supply chain | Locked | | |\n| 17 | 13\u002F31 Conformance engineering | Locked | | |\n\n## Wire evidence\n| Date | Lesson | Mode | Request or scenario | Observed result | Command, cwd, exit |\n|---|---|---|---|---|---|\n\n## Notes\n```\n\nCheck facts that can be observed locally. Ask only for choices or authority\nthat cannot be inferred safely.\n\n## Start Lesson 06 in ten minutes\n\nOn the first invocation, begin the lesson immediately. From the repository\nroot, run:\n\n```bash\npython3 phases\u002F13-tools-and-protocols\u002F06-mcp-fundamentals\u002Fcode\u002Fmain.py\n```\n\nAsk the learner to identify the repeated protocol version and client\ncapabilities, the complete `server\u002Fdiscover` result, error `-32022`, and the\nabsence of protocol-session creation or teardown. Record those observations\nbefore expanding into the rest of Lesson 06.\n\nIf the command cannot run, show one modern request and response from the\nlesson, ask the learner to label every envelope field, and record the result as\nconceptual evidence. Keep the command checkpoint pending.\n\n## Enforce the public deployment gate\n\nBefore any non-loopback bind, shared ingress, hosted endpoint, registry\npublication, or other public deployment, read `publicDeploymentGate` from the\nmanifest. Require the executable Lesson 15 checkpoint, review the target and\nrequested authority, and obtain the learner's explicit confirmation for the\nexternal action.\n\nIf any required evidence is missing, teach or rerun Lesson 15 and keep the\ndeployment action pending. A skill invocation does not grant network,\ncredential, publishing, or deployment authority.\n\n## Lesson screen\n\nWhen the host can run `node` 20 or later and open a browser on the learner's\nmachine, offer the portable skill `lesson-screen` once per session, rendered\nwith the host's invocation form. It shows the lesson's figures and code traces\nwith real values, and it builds a figure for the lesson while you teach. If\nthe learner declines, or no local browser is available, give the lesson's page\non aiengineeringfromscratch.com once. Read screen clicks with its `events`\ncommand before each reply. Screen clicks are hints for you. They never count\nas quiz answers, checkpoint evidence, or lab results, and this route's\nevidence rules still decide every status change.\n\n## Teach one lesson\n\n1. Mark the selected row `In progress`. State its manifest path, duration,\n   group, protocol era, and evidence mode.\n2. Frame one production failure that this lesson prevents. Ask the learner to\n   predict the status, JSON-RPC result, or state transition before explaining\n   it.\n3. Draw one request boundary: producer, transport, consumer, and the exact\n   fields each side validates. Keep protocol state, durable application state,\n   transport state, authorization state, and UI state distinct.\n4. Work through Build It and Use It in small sections. For code, explain one\n   invariant, ask for a prediction, then run or trace the smallest case that\n   can falsify it.\n5. Exercise one success and at least one relevant failure. Prefer exact wire\n   evidence: request id, method, protocol era, headers when applicable, body,\n   status or error code, result type, and terminal state. Keep secret values\n   redacted.\n6. Require every item in the lesson's manifest `checkpointEvidence`. Runtime\n   evidence must come from observed output. Conceptual evidence must name the\n   unexecuted command and remaining uncertainty.\n7. Ask every `post` quiz item one at a time. If the quiz has no staged items,\n   ask all items. Do not reveal `correct`, an answer index, or an explanation\n   before the learner responds. Never put a real answer letter or the answer\n   distribution in a reply hint; use `Reply with one letter: \u003CA|B|C|D>.`\n8. Mark the row `Done` only after the lesson checkpoint and quiz. Append one\n   compact Wire evidence row, add the score to Notes, set the next row to\n   `Next`, and update `Current`.\n\nDo not use passing unit tests as a substitute for the named protocol evidence.\nDo not infer HTTP behavior from an in-process function, authorization from\nauthentication, cancellation from a timeout, or conformance from one SDK.\n\n## Close\n\nEnd with the quiz score, the exact checkpoint evidence recorded, any pending\nruntime or security evidence, and the next manifest lesson. Keep the learner\non this route unless they ask to leave it.\n",{"name":5,"description":7},"main","fe6c657dfb8d02cc898fcee7d4cfbf663aae1416","skills\u002Flearn-mcp",13971,"https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch","https:\u002F\u002Faiengineeringfromscratch.com","https:\u002F\u002Fgithub.com\u002Frohitg00","Engineer | DevRel | GDE | CNCF, AAIF, Claude, Devin, PE Ambassador | Docker Captain | AWS CB | Ex - iii, Cloudera, Solo.io, Cerbos, Oracle, Reliance Jio","2026-11-11 20:45:16",{"source":8,"skillName":5,"commands":69,"prompt":79,"zipAvailable":80},[70,73,76],{"tool":71,"command":72},"npx","npx skills add rohitg00\u002Fai-engineering-from-scratch --skill learn-mcp",{"tool":74,"command":75},"bunx","bunx skills add rohitg00\u002Fai-engineering-from-scratch --skill learn-mcp",{"tool":77,"command":78},"pnpm","pnpm dlx skills add rohitg00\u002Fai-engineering-from-scratch --skill learn-mcp","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Flearn-mcp\u002FSKILL.md",true,[],[83,117,143,169,193,214],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":54},"01M6ERNQB1QMAD1FH5Z5ASYHXZ","write-manual","Writes a long technical manual that explains one subject at one pinned version, from its purpose to each request and response. Every listing comes from an offline capture, every claim traces to a ranked source, and an audit must report TOTAL 0. Use it for \"write a manual\", \"write a 101 manual for \u003Csubject>\", \"manual.json\", \"add a manual section\", \"build a capture kit\", \"audit the manual\", or \"pin a spec version and document it\".","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Fwrite-manual\u002FSKILL.md",[],[90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,34,116],"write","write manual","manual","technical-writing","claim","writing","documentation","audit","offline","add","long","technical","use","document","pin","version","build","manuals","specifications","json","report","kit","listing","request","at","source","section",{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":128,"publishedAt":54},"01M6ERNKG9FZ3WGJ87B975SHW5","start-learning","One-time onboarding for the AI Engineering from Scratch curriculum (542 lessons, 20 phases). Interviews the learner, runs the placement quiz, and writes LEARNING.md, a persistent study plan the learn skill drives. Trigger phrases: \"start learning\", \"set up the course\", \"begin the curriculum\", \"onboard me\", \"create my learning plan\"","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Fstart-learning\u002FSKILL.md",[123,124,127],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":125,"categoryName":126,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000038","人力资源",{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[48,47,129,130,131,132,133,29,51,134,135,136,137,40,138,139,140,141,142,34],"start learning","SET","ai-engineering","time","up","create","plan","me","course","curriculum","study","trigger","interviews","my",{"skillNo":144,"name":145,"nameZh":6,"description":146,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":149,"publishedAt":54},"01M6ERNECV0WVSTXHX8KRF9MBJ","motion-figure","Builds figures that move or respond to the reader, in three tiers: SMIL animations and slider widgets inside lessons, interactive mechanism figures that show the named parts of a real system with a computed verdict, and ordered dither bands in the course palette. Triggers: \"animated figure\", \"add a slider widget\", \"interactive figure\", \"show how this system works\", \"make this diagram move\", \"dither masthead\", \"mechanism figure\", \"figure fence\".","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Fmotion-figure\u002FSKILL.md",[],[150,151,152,153,154,155,156,157,158,159,99,160,161,137,162,163,164,165,44,166,167,168],"motion","motion figure","Animation","show","move","respond","palette","diagram","widgets","how","make","figures","SVG","three","system","triggers","with","animations","lessons",{"skillNo":170,"name":171,"nameZh":6,"description":172,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":173,"categories":174,"keywords":178,"publishedAt":54},"01M6ERNAC0DQ8ER3A99VQBXNNZ","mcpa-certification","AI-native tutor and onboarding workflow for the MCPA (Model Context Protocol Associate) certification in AI Engineering from Scratch. Use when a learner wants to prepare for the MCPA, resume their certification path, learn the next lesson interactively, run and verify practical labs, take the diagnostic or a full mock, or remediate weak exam domains from GitHub with Claude Code, Codex, ChatGPT, Cursor, or another agent.","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Fmcpa-certification\u002FSKILL.md",[175,176,177],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":125,"categoryName":126,"taxonomyType":26,"taxonomyTypeDesc":27},{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[179,180,43,181,46,182,183,29,184,185,53,51,186,187,188,189,190,191,45,192,33],"mcpa certification","certification","practical","domains","exam","workflow","code","native","next","run","take","chatgpt","full","ai-native",{"skillNo":194,"name":195,"nameZh":6,"description":196,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":197,"categories":198,"keywords":200,"publishedAt":54},"01M6ERN5TAAAJ7FE9N86SX6M2M","lesson-screen","Opens a lesson screen in the browser next to the terminal while a course tutor teaches. It shows the lesson's registered figures, code traces with real values, and clickable predictions. It also builds a figure for the lesson on the fly with the course figure skills and any figure skill the host has installed: a blueprint, a mechanism, an isometric machine, or a 3D model. Clicks return to the tutor as events. Trigger phrases: \"open the lesson screen\", \"show me this visually\", \"show this in 3D\", \"draw the architecture\", \"visualize this lesson\", \"build a figure for this lesson\", \"lesson screen\".","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Flesson-screen\u002FSKILL.md",[199],{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[43,201,202,153,203,46,204,205,206,207,208,185,209,210,136,187,211,212,213,161],"screen","lesson screen","3D","open","host","Blueprint","predictions","show me","architecture","browser","terminal","fly","events",{"skillNo":215,"name":29,"nameZh":6,"description":216,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":217,"categories":218,"keywords":222,"publishedAt":54},"01M6ERN1HS8HW1TRJVX5P7SPY0","Interactive lesson tutor for the AI Engineering from Scratch curriculum. Reads LEARNING.md, fetches the next lesson, teaches it section by section in the terminal, quizzes at the end, and records progress. Works cloned or entirely over raw.githubusercontent.com — no setup required. Trigger phrases: \"next lesson\", \"teach me\", \"continue the course\", \"let's learn\", \"resume learning\"","https:\u002F\u002Fgithub.com\u002Frohitg00\u002Fai-engineering-from-scratch\u002Fblob\u002Fmain\u002Fskills\u002Flearn\u002FSKILL.md",[219,220,221],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":125,"categoryName":126,"taxonomyType":26,"taxonomyTypeDesc":27},{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[29,43,131,223,51,136,224,187,211,137,138,140,39,225,47,44,226,38,114,34,116,227],"end","progress","continue","interactive-learning","scratch"]