[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M782P3WJYTTBDPJ161PV1F49":3,"skill-same:01M782P3WJYTTBDPJ161PV1F49":47},{"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":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":46},"01M782P3WJYTTBDPJ161PV1F49","onboard",null,"Onboard a fresh or early scaffold after Blueprint is overlaid by tuning commands, standards, adapters, visibility, and context loading. Use for \u002Fonboard, fresh installation setup, or what to do after installing Blueprint. Use adopt for an established app.","bradtraversy\u002Fskillpass","bradtraversy","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F5550850?v=4","01M782H8BT6A97A17XDFHDABWZ","Brad Traversy","01M782H8CDJZW3HYDHX23FA9VF",26,"MIT","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fonboard\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: onboard\ndescription: Onboard a fresh or early scaffold after Blueprint is overlaid by tuning commands, standards, adapters, visibility, and context loading. Use for \u002Fonboard, fresh installation setup, or what to do after installing Blueprint. Use adopt for an established app.\ndisable-model-invocation: true\n---\n\n# onboard - finish the Blueprint overlay setup\n\n**Context reuse:** Reuse any required file already loaded in project instructions or the current session. Read it again only if absent, changed, or exact current bytes or line references are needed.\n\n**First action:** Before project inspection, preflight, or any other tool call,\npublish `running` to `blueprint\u002F.state\u002Frun.json` using the dashboard activity\ncontract in `AGENTS.md`.\n\nWhere this sits in the workflow:\n\n    scaffold app  ->  overlay Blueprint  ->  [onboard]  ->  project-plan + build-plan  ->  \u002Foverview\n    (user\u002Ftool)       (copied files)          (tune setup)   (user-owned inputs)       (generated context)\n\n`\u002Fonboard` is the fresh-project on-ramp. It assumes the app was scaffolded first\nand the Blueprint files were overlaid after. Run it before filling in plans or\nrunning `\u002Foverview`. Its job is to make the Blueprint fit the real project before\nplanning starts: commands, project title, conventions, ignore rules, and tool\nadapters. It also asks whether the Blueprint workflow files should be committed\nwith the repo or kept local-only through `.gitignore`.\n\nUse `\u002Fadopt` instead when the app is brownfield: real routes, shipped features,\nand project behavior already exist and need to be reflected into the plans.\n\n## Input\n\nNo argument is required. If the user provides context about the stack, hosting,\ndatabase, auth, or preferred tool, use it as a hint and verify against files.\n\n## Step 0 - confirm Git and make an unborn repository usable\n\nBefore reading application code or changing setup files, confirm both Git states:\n\n```bash\ngit rev-parse --is-inside-work-tree\ngit rev-parse --verify HEAD\n```\n\nIf this is not a Git repository, stop and ask the user to initialize one, then\nrerun `\u002Fonboard`.\n\nAn existing first commit may contain only the scaffold or may already contain\nBlueprint. Both are valid. Do not ask the user to rewrite either history shape.\n\nIf Git reports an unborn `HEAD`, handle it here instead of sending the user away\nto run Git commands:\n\n1. Inspect status and build a safe scaffold-only candidate from paths outside\n   `AGENTS.md`, `CLAUDE.md`, `.agents\u002F`, `.claude\u002F`, and `blueprint\u002F`. Exclude\n   secrets, dependencies, caches, build output, generated state, and anything\n   else that should not enter source control. Include the existing `.gitignore`\n   when it is safe.\n2. Resolve the intended default branch from a remote default when available,\n   then an existing `main` or `master`, then Git's configured initial branch,\n   and otherwise `main`. Preserve the current unborn branch name as the setup\n   branch when it is not the intended default. If the intended name is genuinely\n   ambiguous, ask only that one question.\n3. Show the exact candidate and the branch result, then ask once:\n   `Create the initial scaffold commit and continue Onboard? (Recommended)`\n   State that this creates one local commit and never pushes.\n4. On approval, stage only the reviewed candidate, verify the staged diff, and\n   commit it as `chore: scaffold application`. If needed, rename the unborn\n   branch before committing so the root commit establishes the intended default\n   branch. Then create or return to the named setup branch at that same commit\n   and continue Onboard in the same run.\n\nIf there is no safe scaffold candidate, stop with the exact blocker rather than\ncreating an empty or mixed root commit. Never create the commit without explicit\napproval.\n\nThen confirm this is onboarding, not adoption.\n\nInspect the repository and the two planning docs:\n\n- If `blueprint\u002Fproject-plan.md` and `blueprint\u002Fbuild-plan.md` are mostly empty or\n  worksheet-like, proceed.\n- If the app already has substantial shipped features, stop and recommend\n  `\u002Fadopt` instead.\n- If the plans already contain real user-owned content, do not overwrite them.\n  Continue only with setup files such as `AGENTS.md`, `coding-standards.md`,\n  `.gitignore`, and optional notes.\n\nNever run a framework scaffolder. The Blueprint is already overlaid.\n\n## Step 1 - survey the project facts\n\nRead only enough to identify the setup:\n\n- package manager and lockfile (`pnpm-lock.yaml`, `package-lock.json`,\n  `yarn.lock`, `bun.lockb`, etc.)\n- manifest scripts (`package.json`, `pyproject.toml`, `go.mod`, `Cargo.toml`, and\n  similar)\n- framework and runtime config (`astro.config.*`, `next.config.*`, `vite.config.*`,\n  `tailwind.config.*`, database config, test config)\n- source layout, route layout, and app\u002Fpackage directories\n- existing `.gitignore`\n- `blueprint\u002F.state\u002Fmanifest.json`, when present, and its selected adapter list\n- which selected tools need `.agents\u002F`, `.claude\u002F`, or both\n- whether Blueprint workflow paths are already tracked by git\n- existing verification commands and `.github\u002Fworkflows\u002F`\n- `blueprint\u002Fconfig.json`, when present, and whether it parses cleanly\n- project name, from `package.json`, the folder name, existing docs, or the user\n\nDo not infer more than the files support. Mark uncertain items as `> TODO` in the\nsummary rather than inventing a convention.\n\n## Step 2 - update project entry files\n\nIf the root `README.md` is a copied Blueprint workflow document, replace that\nobsolete overlay content in the product README slot:\n\n- Detect it conservatively: the first heading is `# AI Coding Blueprint`, or the\n  opening section clearly describes the Blueprint workflow rather than this app.\n- Create a small root `README.md` stub for the actual project using the detected\n  project name, one-line purpose when known, and the Commands from `AGENTS.md`.\n  Keep it minimal if the project plan is not filled yet.\n- Do not move or copy the workflow document into `blueprint\u002F`. Agents use the\n  local skills, plans, and context files directly.\n- Remove any `AGENTS.md` claim that a project README explains the Blueprint\n  workflow.\n\nIf the root `README.md` already looks like a real project README, leave it alone.\nNever replace a project README with Blueprint documentation.\n\nUpdate the Commands section of `AGENTS.md` to match real scripts and commands.\nRemove the shipped `\u003C!-- blueprint:onboarding-required -->` marker and the `For\na standard Next.js project` instruction when replacing the placeholder\ncommands. Status uses the dedicated marker, with the old sentence retained only\nas a migration fallback, to distinguish a fresh overlay from a tuned project.\nInclude only commands that exist or are intentionally available:\n\n- dev server\n- build\n- preview or start\n- lint, format, typecheck, and test, if configured\n- verify, when a real combined verification command already exists\n- useful app-specific commands, if obvious\n\nIf no test command exists, say so explicitly. Do not claim tests are a gate until\na real test command is configured.\n\nFollow and preserve the proportional-engineering contract in `AGENTS.md` when\nadapting commands, standards, and context to the detected project.\n\nIf `CLAUDE.md` exists and still has the placeholder `# Project Name`, replace it\nwith the detected project name. Keep `@AGENTS.md`. Remove direct imports of\n`project-overview.md`, `current-feature.md`, `coding-standards.md`, and\n`ai-interaction.md`; workflow skills read those files only when relevant.\nPreserve unrelated user imports. Do not move detailed app context into\n`CLAUDE.md`; that belongs in `AGENTS.md` and the generated project overview.\n\n## Step 3 - tune coding standards\n\nUpdate `blueprint\u002Fcontext\u002Fcoding-standards.md` so it matches the detected stack.\nKeep stable, tool-agnostic sections such as writing style, comments, scope, and\ntesting philosophy. Replace stack-specific defaults that do not apply.\n\nCover the practical conventions the build loop needs:\n\n- framework and rendering model\n- package manager\n- project structure\n- styling approach\n- data access and API boundaries, if known\n- validation and error handling expectations\n- test gate status\n- build and verification commands, via `AGENTS.md`\n\nIf the project is too new to reveal a convention, leave a concise `> TODO` rather\nthan pretending a pattern exists.\n\n## Step 4 - check project configuration and AI interaction rules\n\nRead `blueprint\u002Fconfig.json`. A missing file means built-in defaults and is not\nan error. If the file exists but is invalid, stop and show the exact invalid key\nor value before changing other setup files.\n\nKeep project configuration deterministic. Ask before changing preferences and\nedit only values the user actually chose, such as branch prefixes, UI evidence,\nlogic-test strictness, regular or Continuous quality gates, review execution, or\nContinuous Mode limits. Independent review defaults to `when-sensitive` for\nregular and Continuous work, and its execution defaults to `automatic`. Audit,\ncheck, and try guide default to `manual`. Preserve these defaults unless the user\nchooses different policies. A `manual` independent-review gate disables\nautomatic selection for that workflow without disabling explicit independent\naudits. Never put\ncommands, product requirements, communication prose, secrets, or permission for\ncommits, merges, pushes, deployments, publication, destructive actions, failed\nchecks, or finding waivers into config.\n\nUnless the user already chose these values, ask one short **Implementation\nstyle** question using the current tool's selectable prompt when available:\n\n1. **Efficient (Recommended)** - one feature-level review packet, a final code\n   walkthrough option, and no step checkpoint prompts. Write\n   `workflow.stepReview: \"feature\"` and `workflow.checkpointCommits: \"disabled\"`.\n2. **Guided** - pause for approval after every step and offer optional checkpoint\n   commits, followed by the same final code walkthrough option. Write\n   `workflow.stepReview: \"every\"` and `workflow.checkpointCommits: \"enabled\"`.\n3. **Custom** - ask separately when review should happen and whether checkpoint\n   commits should be offered, then write the selected low-level values.\n\nThese are onboarding presets, not a third configuration field. Never write an\n`implementationStyle` key. Show the current two values before asking, preserve\nthem if the user chooses not to change them, and explain that either value can be\nedited later. A later `\u002Fimplement` run reads the current configuration.\nThe final code walkthrough is not a configuration setting and remains available\nwith every implementation style.\n\nRead `blueprint\u002Fcontext\u002Fai-interaction.md` and update only obvious mismatches.\nUsually the default review loop should stay intact. Flag preferences for the user\ninstead of guessing, such as:\n\n- whether review should happen once per feature (the lower-context default) or\n  after every step for teaching, close pairing, or high-risk work\n- whether optional step checkpoint commits should be enabled. Explain that the\n  previous workflow requires per-step review and enabled checkpoints together;\n  changing only `stepReview` restores the approval pauses, not checkpoint prompts\n- whether branches should use a different naming pattern\n- whether `\u002Fcheck` should require browser evidence for UI work\n- whether audit, independent review, check, or try guides should stay manual, run only for their\n  documented conditional case, or run for every regular or Continuous work item\n- whether a selected independent review should use the default automatic\n  isolated reviewer or `review.independentExecution: \"manual\"` for a\n  fresh-session handoff. Explain that automatic execution uses a fresh isolated\n  reviewer only when the active adapter can expose its exact identity and model,\n  and otherwise stops with the manual handoff\n\nIf no changes are needed, say so.\n\n## Step 5 - point to optional CI setup\n\nDo not create or change Verify commands or GitHub workflows during onboarding.\nReport any verification command or CI already present. When equivalent automatic\npull-request checks are absent, mention the optional standalone setup:\n\n```text\nRun \u002Fci or $ci when you want automatic GitHub checks.\n```\n\nExplain that CI is not required to continue with planning or the Blueprint build\nloop. The `\u002Fci` skill owns project-specific Verify and GitHub workflow setup.\n\n## Step 6 - check ignore files, visibility, and adapters\n\nUpdate `.gitignore` for common generated files from the detected stack while\npreserving existing entries. Typical examples include dependencies, build output,\nframework caches, logs, environment files, test output, temporary files, and OS or\neditor files.\n\nAsk how Blueprint workflow files should be handled in git, unless the user\nalready gave a preference:\n\n```text\nBlueprint visibility?\n\n1. Commit Blueprint workflow files\n   Portable. Best for teams and working across machines.\n\n2. Keep Blueprint workflow files local\n   Adds .agents\u002F, .claude\u002F, blueprint\u002F, and CLAUDE.md to .gitignore.\n   Keeps AGENTS.md public as the lightweight project agent guide.\n```\n\nRecommend option 1 by default. If the user chooses option 2:\n\n- Add this block to `.gitignore`, preserving existing entries:\n\n  ```gitignore\n  # AI Blueprint local workflow files\n  .agents\u002F\n  .claude\u002F\n  blueprint\u002F\n  CLAUDE.md\n  ```\n\n- Keep `AGENTS.md` tracked. It remains the lightweight public project guide for\n  commands and conventions.\n- Make `AGENTS.md` public-safe: keep project description, commands, testing gate,\n  and coding conventions, but remove or avoid Blueprint workflow explanations,\n  hidden adapter paths, workflow-document pointers, and core skill lists that\n  would expose the local-only workflow.\n- Explain that local-only mode hides the workflow contents from the repo, but the\n  `.gitignore` names still reveal the ignored paths.\n- Explain that Blueprint state, specs, findings, and history will not travel\n  with the repo; another machine needs the Blueprint reinstalled or restored\n  locally.\n- If any of `.agents\u002F`, `.claude\u002F`, `blueprint\u002F`, or\n  `CLAUDE.md` are already tracked, say `.gitignore` will not hide tracked files.\n  Ask before running\n  `git rm --cached -r .agents .claude blueprint CLAUDE.md`, and\n  only run it if the user explicitly approves. Never delete the local files.\n\nThen report which selected tools and adapter folders are needed:\n\n- When a valid `blueprint\u002F.state\u002Fmanifest.json` exists, its `adapters` list is\n  the authoritative installer selection. The presence of `.agents\u002F` means its\n  files are compatible with Codex, GitHub Copilot, and OpenCode; it does not mean\n  all three tools were selected.\n- Do not ask the user to select adapters again when that valid manifest exists.\n  Keep and report the exact selection. If a required adapter tree is missing,\n  report the mismatch and point to `\u002Fdoctor` instead of guessing or deleting\n  another tree.\n- Without a valid manifest, explain that folder detection cannot distinguish\n  Codex, GitHub Copilot, and OpenCode, then ask which tools the user actually\n  uses instead of assuming all of them are selected.\n- Codex only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`; `CLAUDE.md` and\n  `.claude\u002F` can be deleted.\n- Claude Code only: keep `AGENTS.md`, `CLAUDE.md`, `.claude\u002F`, and `blueprint\u002F`;\n  `.agents\u002F` can be deleted.\n- GitHub Copilot only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`.\n- OpenCode only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`.\n- OpenCode with Claude Code: OpenCode can reuse `.claude\u002F`; no separate\n  `.opencode\u002Fskills\u002F` copy is needed.\n- Mixed tools: keep only the compatible adapter trees required by the selected\n  tools. Never duplicate Blueprint skills under `.opencode\u002Fskills\u002F` because\n  OpenCode already discovers `.agents\u002Fskills\u002F` and `.claude\u002Fskills\u002F`.\n\nDo not delete adapters unless the user explicitly asks.\n\n## Step 7 - hand off to planning\n\nStop with a concise onboarding report:\n\n- stack and package manager detected\n- project name used for entry files\n- README handling, especially if the copied Blueprint README was moved\n- Blueprint visibility choice\n- tracked-file warning if local-only mode was chosen after files were already tracked\n- files changed\n- project configuration state and any user-selected overrides\n- commands now available\n- testing gate status\n- verification command and GitHub checks status\n- adapter recommendation\n- TODOs or uncertainties\n- exact next files for the user to fill in:\n  - `blueprint\u002Fproject-plan.md`\n  - `blueprint\u002Fbuild-plan.md`\n\nMake the direct path clear: the user can write or develop those files through\nany conversation, then run `\u002Foverview`. Also mention `\u002Fdiscovery` or `$discovery`\nas an optional deep planning conversation for users who want guided help. Do not\nstart it, make it a prerequisite, or imply that directly written plans are less\ncomplete.\n\nEnd with the next command:\n\n```text\n\u002Foverview\n```\n\nFor Codex, also mention:\n\n```text\n$overview\n```\n\n## Rules\n\n- Setup files are fair game; planning docs are user-owned.\n- `\u002Fdiscovery` is optional and never runs as part of onboarding. The direct\n  plan-writing path must remain fully supported.\n- Never overwrite real `project-plan.md` or `build-plan.md` content.\n- Never run scaffolders or install dependencies unless the user explicitly asks.\n- Reflect the stack that exists, not the stack the default Blueprint mentions.\n- Be honest about tests. No `test` command means no required test gate yet.\n- Keep `AGENTS.md` public in local-only mode unless the user explicitly asks for\n  a more advanced setup.\n- Do not untrack Blueprint files with `git rm --cached` without a separate\n  explicit approval.\n- Keep changes small and explain what changed.\n\n## Formatting\n\nFormat the output to match the project's conventions in\n`blueprint\u002Fcontext\u002Fai-interaction.md`: concise, scannable markdown, with lists for\nenumerations and tables for matrices rather than dense paragraphs.\n","# onboard - finish the Blueprint overlay setup\n\n**Context reuse:** Reuse any required file already loaded in project instructions or the current session. Read it again only if absent, changed, or exact current bytes or line references are needed.\n\n**First action:** Before project inspection, preflight, or any other tool call,\npublish `running` to `blueprint\u002F.state\u002Frun.json` using the dashboard activity\ncontract in `AGENTS.md`.\n\nWhere this sits in the workflow:\n\n    scaffold app  ->  overlay Blueprint  ->  [onboard]  ->  project-plan + build-plan  ->  \u002Foverview\n    (user\u002Ftool)       (copied files)          (tune setup)   (user-owned inputs)       (generated context)\n\n`\u002Fonboard` is the fresh-project on-ramp. It assumes the app was scaffolded first\nand the Blueprint files were overlaid after. Run it before filling in plans or\nrunning `\u002Foverview`. Its job is to make the Blueprint fit the real project before\nplanning starts: commands, project title, conventions, ignore rules, and tool\nadapters. It also asks whether the Blueprint workflow files should be committed\nwith the repo or kept local-only through `.gitignore`.\n\nUse `\u002Fadopt` instead when the app is brownfield: real routes, shipped features,\nand project behavior already exist and need to be reflected into the plans.\n\n## Input\n\nNo argument is required. If the user provides context about the stack, hosting,\ndatabase, auth, or preferred tool, use it as a hint and verify against files.\n\n## Step 0 - confirm Git and make an unborn repository usable\n\nBefore reading application code or changing setup files, confirm both Git states:\n\n```bash\ngit rev-parse --is-inside-work-tree\ngit rev-parse --verify HEAD\n```\n\nIf this is not a Git repository, stop and ask the user to initialize one, then\nrerun `\u002Fonboard`.\n\nAn existing first commit may contain only the scaffold or may already contain\nBlueprint. Both are valid. Do not ask the user to rewrite either history shape.\n\nIf Git reports an unborn `HEAD`, handle it here instead of sending the user away\nto run Git commands:\n\n1. Inspect status and build a safe scaffold-only candidate from paths outside\n   `AGENTS.md`, `CLAUDE.md`, `.agents\u002F`, `.claude\u002F`, and `blueprint\u002F`. Exclude\n   secrets, dependencies, caches, build output, generated state, and anything\n   else that should not enter source control. Include the existing `.gitignore`\n   when it is safe.\n2. Resolve the intended default branch from a remote default when available,\n   then an existing `main` or `master`, then Git's configured initial branch,\n   and otherwise `main`. Preserve the current unborn branch name as the setup\n   branch when it is not the intended default. If the intended name is genuinely\n   ambiguous, ask only that one question.\n3. Show the exact candidate and the branch result, then ask once:\n   `Create the initial scaffold commit and continue Onboard? (Recommended)`\n   State that this creates one local commit and never pushes.\n4. On approval, stage only the reviewed candidate, verify the staged diff, and\n   commit it as `chore: scaffold application`. If needed, rename the unborn\n   branch before committing so the root commit establishes the intended default\n   branch. Then create or return to the named setup branch at that same commit\n   and continue Onboard in the same run.\n\nIf there is no safe scaffold candidate, stop with the exact blocker rather than\ncreating an empty or mixed root commit. Never create the commit without explicit\napproval.\n\nThen confirm this is onboarding, not adoption.\n\nInspect the repository and the two planning docs:\n\n- If `blueprint\u002Fproject-plan.md` and `blueprint\u002Fbuild-plan.md` are mostly empty or\n  worksheet-like, proceed.\n- If the app already has substantial shipped features, stop and recommend\n  `\u002Fadopt` instead.\n- If the plans already contain real user-owned content, do not overwrite them.\n  Continue only with setup files such as `AGENTS.md`, `coding-standards.md`,\n  `.gitignore`, and optional notes.\n\nNever run a framework scaffolder. The Blueprint is already overlaid.\n\n## Step 1 - survey the project facts\n\nRead only enough to identify the setup:\n\n- package manager and lockfile (`pnpm-lock.yaml`, `package-lock.json`,\n  `yarn.lock`, `bun.lockb`, etc.)\n- manifest scripts (`package.json`, `pyproject.toml`, `go.mod`, `Cargo.toml`, and\n  similar)\n- framework and runtime config (`astro.config.*`, `next.config.*`, `vite.config.*`,\n  `tailwind.config.*`, database config, test config)\n- source layout, route layout, and app\u002Fpackage directories\n- existing `.gitignore`\n- `blueprint\u002F.state\u002Fmanifest.json`, when present, and its selected adapter list\n- which selected tools need `.agents\u002F`, `.claude\u002F`, or both\n- whether Blueprint workflow paths are already tracked by git\n- existing verification commands and `.github\u002Fworkflows\u002F`\n- `blueprint\u002Fconfig.json`, when present, and whether it parses cleanly\n- project name, from `package.json`, the folder name, existing docs, or the user\n\nDo not infer more than the files support. Mark uncertain items as `> TODO` in the\nsummary rather than inventing a convention.\n\n## Step 2 - update project entry files\n\nIf the root `README.md` is a copied Blueprint workflow document, replace that\nobsolete overlay content in the product README slot:\n\n- Detect it conservatively: the first heading is `# AI Coding Blueprint`, or the\n  opening section clearly describes the Blueprint workflow rather than this app.\n- Create a small root `README.md` stub for the actual project using the detected\n  project name, one-line purpose when known, and the Commands from `AGENTS.md`.\n  Keep it minimal if the project plan is not filled yet.\n- Do not move or copy the workflow document into `blueprint\u002F`. Agents use the\n  local skills, plans, and context files directly.\n- Remove any `AGENTS.md` claim that a project README explains the Blueprint\n  workflow.\n\nIf the root `README.md` already looks like a real project README, leave it alone.\nNever replace a project README with Blueprint documentation.\n\nUpdate the Commands section of `AGENTS.md` to match real scripts and commands.\nRemove the shipped `\u003C!-- blueprint:onboarding-required -->` marker and the `For\na standard Next.js project` instruction when replacing the placeholder\ncommands. Status uses the dedicated marker, with the old sentence retained only\nas a migration fallback, to distinguish a fresh overlay from a tuned project.\nInclude only commands that exist or are intentionally available:\n\n- dev server\n- build\n- preview or start\n- lint, format, typecheck, and test, if configured\n- verify, when a real combined verification command already exists\n- useful app-specific commands, if obvious\n\nIf no test command exists, say so explicitly. Do not claim tests are a gate until\na real test command is configured.\n\nFollow and preserve the proportional-engineering contract in `AGENTS.md` when\nadapting commands, standards, and context to the detected project.\n\nIf `CLAUDE.md` exists and still has the placeholder `# Project Name`, replace it\nwith the detected project name. Keep `@AGENTS.md`. Remove direct imports of\n`project-overview.md`, `current-feature.md`, `coding-standards.md`, and\n`ai-interaction.md`; workflow skills read those files only when relevant.\nPreserve unrelated user imports. Do not move detailed app context into\n`CLAUDE.md`; that belongs in `AGENTS.md` and the generated project overview.\n\n## Step 3 - tune coding standards\n\nUpdate `blueprint\u002Fcontext\u002Fcoding-standards.md` so it matches the detected stack.\nKeep stable, tool-agnostic sections such as writing style, comments, scope, and\ntesting philosophy. Replace stack-specific defaults that do not apply.\n\nCover the practical conventions the build loop needs:\n\n- framework and rendering model\n- package manager\n- project structure\n- styling approach\n- data access and API boundaries, if known\n- validation and error handling expectations\n- test gate status\n- build and verification commands, via `AGENTS.md`\n\nIf the project is too new to reveal a convention, leave a concise `> TODO` rather\nthan pretending a pattern exists.\n\n## Step 4 - check project configuration and AI interaction rules\n\nRead `blueprint\u002Fconfig.json`. A missing file means built-in defaults and is not\nan error. If the file exists but is invalid, stop and show the exact invalid key\nor value before changing other setup files.\n\nKeep project configuration deterministic. Ask before changing preferences and\nedit only values the user actually chose, such as branch prefixes, UI evidence,\nlogic-test strictness, regular or Continuous quality gates, review execution, or\nContinuous Mode limits. Independent review defaults to `when-sensitive` for\nregular and Continuous work, and its execution defaults to `automatic`. Audit,\ncheck, and try guide default to `manual`. Preserve these defaults unless the user\nchooses different policies. A `manual` independent-review gate disables\nautomatic selection for that workflow without disabling explicit independent\naudits. Never put\ncommands, product requirements, communication prose, secrets, or permission for\ncommits, merges, pushes, deployments, publication, destructive actions, failed\nchecks, or finding waivers into config.\n\nUnless the user already chose these values, ask one short **Implementation\nstyle** question using the current tool's selectable prompt when available:\n\n1. **Efficient (Recommended)** - one feature-level review packet, a final code\n   walkthrough option, and no step checkpoint prompts. Write\n   `workflow.stepReview: \"feature\"` and `workflow.checkpointCommits: \"disabled\"`.\n2. **Guided** - pause for approval after every step and offer optional checkpoint\n   commits, followed by the same final code walkthrough option. Write\n   `workflow.stepReview: \"every\"` and `workflow.checkpointCommits: \"enabled\"`.\n3. **Custom** - ask separately when review should happen and whether checkpoint\n   commits should be offered, then write the selected low-level values.\n\nThese are onboarding presets, not a third configuration field. Never write an\n`implementationStyle` key. Show the current two values before asking, preserve\nthem if the user chooses not to change them, and explain that either value can be\nedited later. A later `\u002Fimplement` run reads the current configuration.\nThe final code walkthrough is not a configuration setting and remains available\nwith every implementation style.\n\nRead `blueprint\u002Fcontext\u002Fai-interaction.md` and update only obvious mismatches.\nUsually the default review loop should stay intact. Flag preferences for the user\ninstead of guessing, such as:\n\n- whether review should happen once per feature (the lower-context default) or\n  after every step for teaching, close pairing, or high-risk work\n- whether optional step checkpoint commits should be enabled. Explain that the\n  previous workflow requires per-step review and enabled checkpoints together;\n  changing only `stepReview` restores the approval pauses, not checkpoint prompts\n- whether branches should use a different naming pattern\n- whether `\u002Fcheck` should require browser evidence for UI work\n- whether audit, independent review, check, or try guides should stay manual, run only for their\n  documented conditional case, or run for every regular or Continuous work item\n- whether a selected independent review should use the default automatic\n  isolated reviewer or `review.independentExecution: \"manual\"` for a\n  fresh-session handoff. Explain that automatic execution uses a fresh isolated\n  reviewer only when the active adapter can expose its exact identity and model,\n  and otherwise stops with the manual handoff\n\nIf no changes are needed, say so.\n\n## Step 5 - point to optional CI setup\n\nDo not create or change Verify commands or GitHub workflows during onboarding.\nReport any verification command or CI already present. When equivalent automatic\npull-request checks are absent, mention the optional standalone setup:\n\n```text\nRun \u002Fci or $ci when you want automatic GitHub checks.\n```\n\nExplain that CI is not required to continue with planning or the Blueprint build\nloop. The `\u002Fci` skill owns project-specific Verify and GitHub workflow setup.\n\n## Step 6 - check ignore files, visibility, and adapters\n\nUpdate `.gitignore` for common generated files from the detected stack while\npreserving existing entries. Typical examples include dependencies, build output,\nframework caches, logs, environment files, test output, temporary files, and OS or\neditor files.\n\nAsk how Blueprint workflow files should be handled in git, unless the user\nalready gave a preference:\n\n```text\nBlueprint visibility?\n\n1. Commit Blueprint workflow files\n   Portable. Best for teams and working across machines.\n\n2. Keep Blueprint workflow files local\n   Adds .agents\u002F, .claude\u002F, blueprint\u002F, and CLAUDE.md to .gitignore.\n   Keeps AGENTS.md public as the lightweight project agent guide.\n```\n\nRecommend option 1 by default. If the user chooses option 2:\n\n- Add this block to `.gitignore`, preserving existing entries:\n\n  ```gitignore\n  # AI Blueprint local workflow files\n  .agents\u002F\n  .claude\u002F\n  blueprint\u002F\n  CLAUDE.md\n  ```\n\n- Keep `AGENTS.md` tracked. It remains the lightweight public project guide for\n  commands and conventions.\n- Make `AGENTS.md` public-safe: keep project description, commands, testing gate,\n  and coding conventions, but remove or avoid Blueprint workflow explanations,\n  hidden adapter paths, workflow-document pointers, and core skill lists that\n  would expose the local-only workflow.\n- Explain that local-only mode hides the workflow contents from the repo, but the\n  `.gitignore` names still reveal the ignored paths.\n- Explain that Blueprint state, specs, findings, and history will not travel\n  with the repo; another machine needs the Blueprint reinstalled or restored\n  locally.\n- If any of `.agents\u002F`, `.claude\u002F`, `blueprint\u002F`, or\n  `CLAUDE.md` are already tracked, say `.gitignore` will not hide tracked files.\n  Ask before running\n  `git rm --cached -r .agents .claude blueprint CLAUDE.md`, and\n  only run it if the user explicitly approves. Never delete the local files.\n\nThen report which selected tools and adapter folders are needed:\n\n- When a valid `blueprint\u002F.state\u002Fmanifest.json` exists, its `adapters` list is\n  the authoritative installer selection. The presence of `.agents\u002F` means its\n  files are compatible with Codex, GitHub Copilot, and OpenCode; it does not mean\n  all three tools were selected.\n- Do not ask the user to select adapters again when that valid manifest exists.\n  Keep and report the exact selection. If a required adapter tree is missing,\n  report the mismatch and point to `\u002Fdoctor` instead of guessing or deleting\n  another tree.\n- Without a valid manifest, explain that folder detection cannot distinguish\n  Codex, GitHub Copilot, and OpenCode, then ask which tools the user actually\n  uses instead of assuming all of them are selected.\n- Codex only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`; `CLAUDE.md` and\n  `.claude\u002F` can be deleted.\n- Claude Code only: keep `AGENTS.md`, `CLAUDE.md`, `.claude\u002F`, and `blueprint\u002F`;\n  `.agents\u002F` can be deleted.\n- GitHub Copilot only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`.\n- OpenCode only: keep `AGENTS.md`, `.agents\u002F`, and `blueprint\u002F`.\n- OpenCode with Claude Code: OpenCode can reuse `.claude\u002F`; no separate\n  `.opencode\u002Fskills\u002F` copy is needed.\n- Mixed tools: keep only the compatible adapter trees required by the selected\n  tools. Never duplicate Blueprint skills under `.opencode\u002Fskills\u002F` because\n  OpenCode already discovers `.agents\u002Fskills\u002F` and `.claude\u002Fskills\u002F`.\n\nDo not delete adapters unless the user explicitly asks.\n\n## Step 7 - hand off to planning\n\nStop with a concise onboarding report:\n\n- stack and package manager detected\n- project name used for entry files\n- README handling, especially if the copied Blueprint README was moved\n- Blueprint visibility choice\n- tracked-file warning if local-only mode was chosen after files were already tracked\n- files changed\n- project configuration state and any user-selected overrides\n- commands now available\n- testing gate status\n- verification command and GitHub checks status\n- adapter recommendation\n- TODOs or uncertainties\n- exact next files for the user to fill in:\n  - `blueprint\u002Fproject-plan.md`\n  - `blueprint\u002Fbuild-plan.md`\n\nMake the direct path clear: the user can write or develop those files through\nany conversation, then run `\u002Foverview`. Also mention `\u002Fdiscovery` or `$discovery`\nas an optional deep planning conversation for users who want guided help. Do not\nstart it, make it a prerequisite, or imply that directly written plans are less\ncomplete.\n\nEnd with the next command:\n\n```text\n\u002Foverview\n```\n\nFor Codex, also mention:\n\n```text\n$overview\n```\n\n## Rules\n\n- Setup files are fair game; planning docs are user-owned.\n- `\u002Fdiscovery` is optional and never runs as part of onboarding. The direct\n  plan-writing path must remain fully supported.\n- Never overwrite real `project-plan.md` or `build-plan.md` content.\n- Never run scaffolders or install dependencies unless the user explicitly asks.\n- Reflect the stack that exists, not the stack the default Blueprint mentions.\n- Be honest about tests. No `test` command means no required test gate yet.\n- Keep `AGENTS.md` public in local-only mode unless the user explicitly asks for\n  a more advanced setup.\n- Do not untrack Blueprint files with `git rm --cached` without a separate\n  explicit approval.\n- Keep changes small and explain what changed.\n\n## Formatting\n\nFormat the output to match the project's conventions in\n`blueprint\u002Fcontext\u002Fai-interaction.md`: concise, scannable markdown, with lists for\nenumerations and tables for matrices rather than dense paragraphs.\n",{"name":5,"description":7,"disable-model-invocation":24},true,"main","0007775402258962018a93be4dbf1cbcbc1536d8",".claude\u002Fskills\u002Fonboard",3,"https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass","https:\u002F\u002Fgithub.com\u002Fbradtraversy","traversymedia.com","Full stack web developer and online instructor at https:\u002F\u002Ftraversymedia.com and https:\u002F\u002Fstart.dev","2026-10-06 00:26:26",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":24},[36,39,42],{"tool":37,"command":38},"npx","npx skills add bradtraversy\u002Fskillpass --skill onboard",{"tool":40,"command":41},"bunx","bunx skills add bradtraversy\u002Fskillpass --skill onboard",{"tool":43,"command":44},"pnpm","pnpm dlx skills add bradtraversy\u002Fskillpass --skill onboard","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fonboard\u002FSKILL.md",[],[48,54,60,65,71,77],{"skillNo":49,"name":50,"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":51,"categories":52,"keywords":53,"publishedAt":19},"01M782QBBKH3WTBZDE9DP52XH4","plan","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fworkflow-pack\u002Fskills\u002Fplan\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"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":57,"categories":58,"keywords":59,"publishedAt":19},"01M782Q980Y1KP1J77TNJHD5H8","apply","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fworkflow-pack\u002Fskills\u002Fapply\u002FSKILL.md",[],[],{"skillNo":61,"name":50,"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":62,"categories":63,"keywords":64,"publishedAt":19},"01M782Q75AS3P8SW7D819TACVD","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fworkflow-pack\u002F.claude\u002Fskills\u002Fplan\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"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":68,"categories":69,"keywords":70,"publishedAt":19},"01M782Q5WAP9DQH14TEKG5ZFX5","undeclared-network","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fundeclared-network\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"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":74,"categories":75,"keywords":76,"publishedAt":19},"01M782Q424YF4TAYYX9TCCZCS8","prompt-injection","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fprompt-injection\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"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":80,"categories":81,"keywords":82,"publishedAt":19},"01M782Q28RQV6MJM98J6EN7XFF","missing-manifest","https:\u002F\u002Fgithub.com\u002Fbradtraversy\u002Fskillpass\u002Fblob\u002Fmain\u002Fpackages\u002Fvalidator\u002Ffixtures\u002Fmissing-manifest\u002FSKILL.md",[],[]]