[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AAZFZQZB1J9YNETWKJY69T":3,"skill-same:01M7AAZFZQZB1J9YNETWKJY69T":53},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":13,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M7AAZFZQZB1J9YNETWKJY69T","blueprint-import",null,"Migrate an existing Home Assistant custom integration into the blueprint's structure and tooling without breaking the installations already running it. Use when asked to \"import my integration\", \"migrate my existing integration into the blueprint\", \"adopt the blueprint in my repository\", \"move my custom component over\", \"restructure my integration to match this layout\", or when switching from another template such as the upstream ludeeus one. Covers what must stay byte-identical, the git and repository strategy, the phase order that keeps every step revertable, the stock-taking audit that produces the backlog, and how to apply this project's rules to legacy code. SYMPTOMS — load this if you are about to: change the integration's domain during the move; run the whole migration as one commit; restructure packages before the imported code is green; let an automated rename touch unique IDs or translation keys; let the import become a rewrite; or start editing without a record of the current entity IDs.","amg0\u002Fha_ksenia","amg0","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F16444411?v=4","01M7AAZDRNF8BXHM6GAN9QWJ41","01M7AAZDSEHKPEWK6PKK22S2NN",0,"MIT","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fblueprint-import\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: blueprint-import\ndescription: >-\n  Migrate an existing Home Assistant custom integration into the blueprint's structure and tooling without\n  breaking the installations already running it. Use when asked to \"import my integration\", \"migrate my existing\n  integration into the blueprint\", \"adopt the blueprint in my repository\", \"move my custom component over\",\n  \"restructure my integration to match this layout\", or when switching from another template such as the upstream\n  ludeeus one. Covers what must stay byte-identical, the git and repository strategy, the phase order that keeps\n  every step revertable, the stock-taking audit that produces the backlog, and how to apply this project's rules\n  to legacy code. SYMPTOMS — load this if you are about to: change the integration's domain during the move; run\n  the whole migration as one commit; restructure packages before the imported code is green; let an automated\n  rename touch unique IDs or translation keys; let the import become a rewrite; or start editing without a\n  record of the current entity IDs.\n---\n\n# Import an existing integration\n\nA published integration has users. Their automations, dashboards, and statistics are bound to entity IDs, unique IDs,\nstate values and action names — none of which the migration is allowed to change. Everything else is negotiable.\n\n**The whole procedure is one long behaviour-preserving refactor.** Treat any observable difference as a bug in the\nmigration, not as an improvement, unless the developer explicitly asked for it.\n\n## Phase 0 — Baseline first, before touching anything\n\nWithout a record of the current behaviour, nothing in the later phases is verifiable.\n\nTwo records are needed, and they are not the same thing. This phase captures what the code **does**.\n[`ha-grill`](..\u002Fha-grill\u002FSKILL.md) captures what the maintainer **wants** — how many installations are at stake,\nwhat is deliberately unusual in there, what they already know is broken, and whether a breaking release is on the\ntable at all. Run that interview first: its answers decide how much of phases 3 to 7 is affordable, and an import\nplanned without them optimises for the wrong constraint.\n\n1. **Capture the identity of every entity.** From a running installation, export the entity IDs and unique IDs of the\n   integration's entities (the entity registry, or the diagnostics download). If no live instance is available, derive\n   them from the code and say so — a derived baseline is weaker.\n2. **Record the contract** in a scratch file under `.agents\u002Fscratch\u002F`:\n   - the domain\n   - the `unique_id` format for the config entry and for each entity\n   - the keys stored in `entry.data` and `entry.options`, and the config entry `VERSION` \u002F `MINOR_VERSION`\n   - the service action names and their field names\n   - the `device_class`, `state_class`, unit and attribute names of every entity\n   - any `Store` \u002F `.storage` key the integration writes\n3. **Note what the integration depends on**: `manifest.json` → `requirements`, the minimum Home Assistant version, and\n   whether it uses discovery.\n\nThis file is the acceptance test for phases 4 to 6. Nothing in it may change without an explicit decision under\n[`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md).\n\n## Phase 1 — Repository, then `initialize.sh`\n\n**Stay in the existing GitHub repository.** Issues, stars, releases and the catalog listing are tied to the repository\nURL, not to the commit history inside it. Pull the blueprint into that repository rather than starting a new one from\nthe template. The git strategies — clean force-push versus merging unrelated histories — and the repository settings\nto re-check are in [`references\u002Fgit-strategy.md`](references\u002Fgit-strategy.md).\n\nThen run `initialize.sh` **with the identifiers the integration already has**:\n\n- `--domain` **must** be the existing domain. It is the primary key of every config entry and the prefix of every\n  entity ID. Changing it does not migrate anything — it orphans every installation. If the developer wants a\n  different domain, stop and treat it as its own breaking change with its own decision.\n- `--title` should match the name users already see, unless a rename is deliberate.\n- `--namespace` sets the class prefix. Picking the prefix the code already uses means no rename happens at all, which\n  is the safest option. If you do pick a new one, phase 2's verification exists precisely to catch what the rename\n  swept up.\n\n`initialize.sh` normally refuses to run outside a fresh single-commit template clone; `--force` is required here. Use\n`--dry-run` first.\n\n## Phase 2 — Drop the code in unchanged, and get it green\n\nCopy `custom_components\u002F\u003Cdomain>\u002F` from the old repository over the blueprint's generated one — **replacing it, not\nmerging into it**. The blueprint's example platforms, API client and entities are demonstrations; keeping any of them\nalongside real code ships broken entities.\n\nHome Assistant does not care about the internal file layout, so a flat set of modules is perfectly valid at this\npoint. Do not restructure yet.\n\nThen make it pass, changing as little as possible:\n\n```bash\nscript\u002Flint          # fixes formatting, reports the rest\nscript\u002Ftype-check\nscript\u002Fhassfest\nscript\u002Ftest\n```\n\nReconcile `manifest.json` → `requirements` with the root `requirements.txt`, and restore any `hacs.json` fields the\nold repository had that `initialize.sh` did not carry over ([`blueprint-tooling`](..\u002Fblueprint-tooling\u002FSKILL.md)).\n\n**Verify against phase 0 and commit.** This checkpoint — old code, new tooling, everything green — is the fallback\nevery later phase can be reverted to. Reaching it is the milestone worth reporting.\n\n## Phase 3 — Take stock, before deciding how much to fix\n\nThe phase 2 checkpoint is the only moment where the whole integration is in front of you, running, with nothing\nchanged yet. Audit it here and the result is an inventory of what you inherited; audit it at the end and it is a\nverdict on your own work.\n\nRun a full audit with [`ha-quality-review`](..\u002Fha-quality-review\u002FSKILL.md) against the imported code. Then sort every\nfinding — the sorting, not the audit, is what makes the result usable:\n\n| Bucket       | What lands in it                                                                                                   | What happens to it                                      |\n| ------------ | ------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------- |\n| **Free**     | Nothing a user can observe: layering, typing, file structure, deprecated APIs, missing tests, absent diagnostics   | Folded into phases 4 to 6 as you reach them             |\n| **Costly**   | Anything on the phase 0 contract: entity names, `state_class`, units, unique ID format, renamed or removed options | Batched — one decision, one migration, one release      |\n| **Deferred** | Needs a real device, a vendor answer, or a decision nobody can make today                                          | Written down with what it blocks, never quietly dropped |\n\n**Costly findings are batched, not dripped.** Five releases that each rename a few entities cost users five\ninterruptions; one release that renames them all, with a migration and a changelog entry they can act on, costs one.\nCollect them, present them together, and let the developer choose the release\n([`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md)).\n\nTurn the result into a phased plan in `.agents\u002Fscratch\u002F` ([`ha-planning`](..\u002Fha-planning\u002FSKILL.md)) — phases 4 to 7\nthen execute it. Weigh it against what the interview established: an integration with thousands of installations whose\nmaintainer does not want a major release has a Free-only backlog, and that is a legitimate outcome, not a failed\nimport.\n\n**This phase changes no code.** A finding you fixed while cataloguing is a finding nobody reviewed, and it spends the\nphase 2 checkpoint you may still need to fall back to.\n\n## Phase 4 — Modernise the APIs, and nothing else\n\nOld integrations accumulate APIs that Home Assistant has since deprecated or removed. Fix those now, while the file\nlayout is still familiar and a diff is easy to read. [`ha-modern-apis`](..\u002Fha-modern-apis\u002FSKILL.md) carries the full\ntable; the ones that show up in nearly every legacy integration, together with the upstream-template patterns worth\nrecognising, are in [`references\u002Flegacy-patterns.md`](references\u002Flegacy-patterns.md).\n\nThis phase is strictly behaviour-preserving. Resist every improvement that changes what a user observes — note it for\nlater instead. Commit separately from phase 5.\n\n## Phase 5 — Restructure into the blueprint layout\n\nOnly now move code into `api\u002F`, `coordinator\u002F`, `entity\u002F`, `config_flow_handler\u002F`, `\u003Cplatform>\u002F`, `service_actions\u002F`,\n`utils\u002F` as described in `docs\u002Fdevelopment\u002FARCHITECTURE.md` and the \"Integration Structure\" section of\n[`AGENTS.md`](..\u002F..\u002F..\u002FAGENTS.md).\n\nThis is file movement and import rewriting, not redesign. Two things to be deliberate about:\n\n- The blueprint keeps a thin `config_flow.py` shim at the top level and the real flow in `config_flow_handler\u002F`.\n  Home Assistant discovers the flow through the shim, so it must stay.\n- One entity class per file. Splitting a large platform module is the main real work of this phase.\n\nAfter moving, re-run the full validation and re-verify the phase 0 contract. Unique IDs and translation keys are the\nthings most easily lost in a bulk rename.\n\n## Phase 6 — Apply this project's rules to the imported code\n\nThe linters are mandatory but they only cover mechanics. What makes the import worthwhile is the guidance the linters\ncannot express — apply it as a review pass, file type by file type, using the matching instructions file and skill:\n\n| Area                      | What typically needs fixing in imported code                                                             |\n| ------------------------- | -------------------------------------------------------------------------------------------------------- |\n| Entities                  | `name=` and `icon=` in code instead of `translation_key` and `icons.json`; missing `state_class`         |\n| Entity\u002Fcoordinator layers | entities calling the API client directly instead of reading `coordinator.data`                           |\n| Service actions           | registered in `async_setup_entry()` instead of `async_setup()`; `services.yaml` fields without selectors |\n| Config flow               | no `unique_id`, or one derived from a host or IP; credentials in `entry.options`                         |\n| Errors                    | bare `HomeAssistantError` with an English message instead of a translated one                            |\n| Diagnostics               | absent, or present without `async_redact_data()`                                                         |\n| Device registry           | unscoped registry lookups, or several subentries sharing one device                                      |\n\n**Each of these can be a breaking change.** Adding a `translation_key` to an entity that had `name=` changes its\nname, and adding a missing `state_class` changes its statistics. Check every one against the phase 0 contract and\nraise it with the developer before implementing — see [`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md).\n\nThe phase 3 backlog says which of these this repository actually has, and which bucket each one fell into — work from\nthat. The table above is what to look for when no audit was run.\n\nWork through this in small commits, not one sweep. If the developer wants only the mandatory parts, phases 0 to 5\nalready stand on their own.\n\n## Phase 7 — Tests, review, release\n\n- Add tests for whatever the old repository had none of, starting with the config flow and one platform\n  ([`ha-testing`](..\u002Fha-testing\u002FSKILL.md)).\n- Re-run [`ha-quality-review`](..\u002Fha-quality-review\u002FSKILL.md). It now measures what the import produced rather than\n  what it inherited, so compare it against the phase 3 audit and say what moved.\n- Start Home Assistant with the **existing** `config\u002F.storage` and confirm the entry loads, the entity IDs are\n  unchanged and history is intact. A mocked suite cannot prove this.\n- Release under [`ha-release`](..\u002Fha-release\u002FSKILL.md). If anything in the phase 0 contract did change, it needs a\n  `BREAKING CHANGE:` footer and a migration note for users.\n\nReport which verification actually happened and which did not ([`AI_POLICY.md`](..\u002F..\u002F..\u002FAI_POLICY.md)).\n\nThe first new feature after an import is the one most likely to be built on assumptions the old repository never\nstated. Run [`ha-grill`](..\u002Fha-grill\u002FSKILL.md) before it, not after.\n\n## Retire this skill when the import is done\n\nImporting happens once. Afterwards this skill is dead weight in every session:\n\n1. Delete `.agents\u002Fskills\u002Fblueprint-import\u002F`.\n2. Remove its rows from the catalogues in `.agents\u002Fskills\u002FREADME.md` and `AGENTS.md`.\n3. Add `.agents\u002Fskills\u002Fblueprint-import\u002F` to `.templatesyncignore`, or the next weekly template-sync pull request\n   restores it.\n\n## Do not\n\n- Do not change the domain, and do not \"tidy up\" a unique ID format, entity ID, or action name in passing.\n- Do not apply the phase 6 rules as routine cleanup: swapping `name=` for a `translation_key` renames the entity in\n  every dashboard, and adding a missing `state_class` rewrites its long-term statistics. Both need approval.\n- Do not run phases 2 to 6 as one commit — a failure then has no revertable checkpoint.\n- Do not let the phase 3 audit turn into a rewrite. It produces a backlog the developer decides on; an import that\n  quietly became a redesign has no checkpoint left and no reviewer who can still tell the two apart.\n- Do not restructure before the imported code is green in its original layout.\n- Do not merge the blueprint's example entities with the imported ones.\n- Do not trust an automated class rename around `unique_id`, `translation_key`, storage keys, or anything else the\n  user's installation persisted; diff them explicitly.\n- Do not abandon the old repository for a fresh one created from the template.\n","# Import an existing integration\n\nA published integration has users. Their automations, dashboards, and statistics are bound to entity IDs, unique IDs,\nstate values and action names — none of which the migration is allowed to change. Everything else is negotiable.\n\n**The whole procedure is one long behaviour-preserving refactor.** Treat any observable difference as a bug in the\nmigration, not as an improvement, unless the developer explicitly asked for it.\n\n## Phase 0 — Baseline first, before touching anything\n\nWithout a record of the current behaviour, nothing in the later phases is verifiable.\n\nTwo records are needed, and they are not the same thing. This phase captures what the code **does**.\n[`ha-grill`](..\u002Fha-grill\u002FSKILL.md) captures what the maintainer **wants** — how many installations are at stake,\nwhat is deliberately unusual in there, what they already know is broken, and whether a breaking release is on the\ntable at all. Run that interview first: its answers decide how much of phases 3 to 7 is affordable, and an import\nplanned without them optimises for the wrong constraint.\n\n1. **Capture the identity of every entity.** From a running installation, export the entity IDs and unique IDs of the\n   integration's entities (the entity registry, or the diagnostics download). If no live instance is available, derive\n   them from the code and say so — a derived baseline is weaker.\n2. **Record the contract** in a scratch file under `.agents\u002Fscratch\u002F`:\n   - the domain\n   - the `unique_id` format for the config entry and for each entity\n   - the keys stored in `entry.data` and `entry.options`, and the config entry `VERSION` \u002F `MINOR_VERSION`\n   - the service action names and their field names\n   - the `device_class`, `state_class`, unit and attribute names of every entity\n   - any `Store` \u002F `.storage` key the integration writes\n3. **Note what the integration depends on**: `manifest.json` → `requirements`, the minimum Home Assistant version, and\n   whether it uses discovery.\n\nThis file is the acceptance test for phases 4 to 6. Nothing in it may change without an explicit decision under\n[`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md).\n\n## Phase 1 — Repository, then `initialize.sh`\n\n**Stay in the existing GitHub repository.** Issues, stars, releases and the catalog listing are tied to the repository\nURL, not to the commit history inside it. Pull the blueprint into that repository rather than starting a new one from\nthe template. The git strategies — clean force-push versus merging unrelated histories — and the repository settings\nto re-check are in [`references\u002Fgit-strategy.md`](references\u002Fgit-strategy.md).\n\nThen run `initialize.sh` **with the identifiers the integration already has**:\n\n- `--domain` **must** be the existing domain. It is the primary key of every config entry and the prefix of every\n  entity ID. Changing it does not migrate anything — it orphans every installation. If the developer wants a\n  different domain, stop and treat it as its own breaking change with its own decision.\n- `--title` should match the name users already see, unless a rename is deliberate.\n- `--namespace` sets the class prefix. Picking the prefix the code already uses means no rename happens at all, which\n  is the safest option. If you do pick a new one, phase 2's verification exists precisely to catch what the rename\n  swept up.\n\n`initialize.sh` normally refuses to run outside a fresh single-commit template clone; `--force` is required here. Use\n`--dry-run` first.\n\n## Phase 2 — Drop the code in unchanged, and get it green\n\nCopy `custom_components\u002F\u003Cdomain>\u002F` from the old repository over the blueprint's generated one — **replacing it, not\nmerging into it**. The blueprint's example platforms, API client and entities are demonstrations; keeping any of them\nalongside real code ships broken entities.\n\nHome Assistant does not care about the internal file layout, so a flat set of modules is perfectly valid at this\npoint. Do not restructure yet.\n\nThen make it pass, changing as little as possible:\n\n```bash\nscript\u002Flint          # fixes formatting, reports the rest\nscript\u002Ftype-check\nscript\u002Fhassfest\nscript\u002Ftest\n```\n\nReconcile `manifest.json` → `requirements` with the root `requirements.txt`, and restore any `hacs.json` fields the\nold repository had that `initialize.sh` did not carry over ([`blueprint-tooling`](..\u002Fblueprint-tooling\u002FSKILL.md)).\n\n**Verify against phase 0 and commit.** This checkpoint — old code, new tooling, everything green — is the fallback\nevery later phase can be reverted to. Reaching it is the milestone worth reporting.\n\n## Phase 3 — Take stock, before deciding how much to fix\n\nThe phase 2 checkpoint is the only moment where the whole integration is in front of you, running, with nothing\nchanged yet. Audit it here and the result is an inventory of what you inherited; audit it at the end and it is a\nverdict on your own work.\n\nRun a full audit with [`ha-quality-review`](..\u002Fha-quality-review\u002FSKILL.md) against the imported code. Then sort every\nfinding — the sorting, not the audit, is what makes the result usable:\n\n| Bucket       | What lands in it                                                                                                   | What happens to it                                      |\n| ------------ | ------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------- |\n| **Free**     | Nothing a user can observe: layering, typing, file structure, deprecated APIs, missing tests, absent diagnostics   | Folded into phases 4 to 6 as you reach them             |\n| **Costly**   | Anything on the phase 0 contract: entity names, `state_class`, units, unique ID format, renamed or removed options | Batched — one decision, one migration, one release      |\n| **Deferred** | Needs a real device, a vendor answer, or a decision nobody can make today                                          | Written down with what it blocks, never quietly dropped |\n\n**Costly findings are batched, not dripped.** Five releases that each rename a few entities cost users five\ninterruptions; one release that renames them all, with a migration and a changelog entry they can act on, costs one.\nCollect them, present them together, and let the developer choose the release\n([`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md)).\n\nTurn the result into a phased plan in `.agents\u002Fscratch\u002F` ([`ha-planning`](..\u002Fha-planning\u002FSKILL.md)) — phases 4 to 7\nthen execute it. Weigh it against what the interview established: an integration with thousands of installations whose\nmaintainer does not want a major release has a Free-only backlog, and that is a legitimate outcome, not a failed\nimport.\n\n**This phase changes no code.** A finding you fixed while cataloguing is a finding nobody reviewed, and it spends the\nphase 2 checkpoint you may still need to fall back to.\n\n## Phase 4 — Modernise the APIs, and nothing else\n\nOld integrations accumulate APIs that Home Assistant has since deprecated or removed. Fix those now, while the file\nlayout is still familiar and a diff is easy to read. [`ha-modern-apis`](..\u002Fha-modern-apis\u002FSKILL.md) carries the full\ntable; the ones that show up in nearly every legacy integration, together with the upstream-template patterns worth\nrecognising, are in [`references\u002Flegacy-patterns.md`](references\u002Flegacy-patterns.md).\n\nThis phase is strictly behaviour-preserving. Resist every improvement that changes what a user observes — note it for\nlater instead. Commit separately from phase 5.\n\n## Phase 5 — Restructure into the blueprint layout\n\nOnly now move code into `api\u002F`, `coordinator\u002F`, `entity\u002F`, `config_flow_handler\u002F`, `\u003Cplatform>\u002F`, `service_actions\u002F`,\n`utils\u002F` as described in `docs\u002Fdevelopment\u002FARCHITECTURE.md` and the \"Integration Structure\" section of\n[`AGENTS.md`](..\u002F..\u002F..\u002FAGENTS.md).\n\nThis is file movement and import rewriting, not redesign. Two things to be deliberate about:\n\n- The blueprint keeps a thin `config_flow.py` shim at the top level and the real flow in `config_flow_handler\u002F`.\n  Home Assistant discovers the flow through the shim, so it must stay.\n- One entity class per file. Splitting a large platform module is the main real work of this phase.\n\nAfter moving, re-run the full validation and re-verify the phase 0 contract. Unique IDs and translation keys are the\nthings most easily lost in a bulk rename.\n\n## Phase 6 — Apply this project's rules to the imported code\n\nThe linters are mandatory but they only cover mechanics. What makes the import worthwhile is the guidance the linters\ncannot express — apply it as a review pass, file type by file type, using the matching instructions file and skill:\n\n| Area                      | What typically needs fixing in imported code                                                             |\n| ------------------------- | -------------------------------------------------------------------------------------------------------- |\n| Entities                  | `name=` and `icon=` in code instead of `translation_key` and `icons.json`; missing `state_class`         |\n| Entity\u002Fcoordinator layers | entities calling the API client directly instead of reading `coordinator.data`                           |\n| Service actions           | registered in `async_setup_entry()` instead of `async_setup()`; `services.yaml` fields without selectors |\n| Config flow               | no `unique_id`, or one derived from a host or IP; credentials in `entry.options`                         |\n| Errors                    | bare `HomeAssistantError` with an English message instead of a translated one                            |\n| Diagnostics               | absent, or present without `async_redact_data()`                                                         |\n| Device registry           | unscoped registry lookups, or several subentries sharing one device                                      |\n\n**Each of these can be a breaking change.** Adding a `translation_key` to an entity that had `name=` changes its\nname, and adding a missing `state_class` changes its statistics. Check every one against the phase 0 contract and\nraise it with the developer before implementing — see [`ha-breaking-changes`](..\u002Fha-breaking-changes\u002FSKILL.md).\n\nThe phase 3 backlog says which of these this repository actually has, and which bucket each one fell into — work from\nthat. The table above is what to look for when no audit was run.\n\nWork through this in small commits, not one sweep. If the developer wants only the mandatory parts, phases 0 to 5\nalready stand on their own.\n\n## Phase 7 — Tests, review, release\n\n- Add tests for whatever the old repository had none of, starting with the config flow and one platform\n  ([`ha-testing`](..\u002Fha-testing\u002FSKILL.md)).\n- Re-run [`ha-quality-review`](..\u002Fha-quality-review\u002FSKILL.md). It now measures what the import produced rather than\n  what it inherited, so compare it against the phase 3 audit and say what moved.\n- Start Home Assistant with the **existing** `config\u002F.storage` and confirm the entry loads, the entity IDs are\n  unchanged and history is intact. A mocked suite cannot prove this.\n- Release under [`ha-release`](..\u002Fha-release\u002FSKILL.md). If anything in the phase 0 contract did change, it needs a\n  `BREAKING CHANGE:` footer and a migration note for users.\n\nReport which verification actually happened and which did not ([`AI_POLICY.md`](..\u002F..\u002F..\u002FAI_POLICY.md)).\n\nThe first new feature after an import is the one most likely to be built on assumptions the old repository never\nstated. Run [`ha-grill`](..\u002Fha-grill\u002FSKILL.md) before it, not after.\n\n## Retire this skill when the import is done\n\nImporting happens once. Afterwards this skill is dead weight in every session:\n\n1. Delete `.agents\u002Fskills\u002Fblueprint-import\u002F`.\n2. Remove its rows from the catalogues in `.agents\u002Fskills\u002FREADME.md` and `AGENTS.md`.\n3. Add `.agents\u002Fskills\u002Fblueprint-import\u002F` to `.templatesyncignore`, or the next weekly template-sync pull request\n   restores it.\n\n## Do not\n\n- Do not change the domain, and do not \"tidy up\" a unique ID format, entity ID, or action name in passing.\n- Do not apply the phase 6 rules as routine cleanup: swapping `name=` for a `translation_key` renames the entity in\n  every dashboard, and adding a missing `state_class` rewrites its long-term statistics. Both need approval.\n- Do not run phases 2 to 6 as one commit — a failure then has no revertable checkpoint.\n- Do not let the phase 3 audit turn into a rewrite. It produces a backlog the developer decides on; an import that\n  quietly became a redesign has no checkpoint left and no reviewer who can still tell the two apart.\n- Do not restructure before the imported code is green in its original layout.\n- Do not merge the blueprint's example entities with the imported ones.\n- Do not trust an automated class rename around `unique_id`, `translation_key`, storage keys, or anything else the\n  user's installation persisted; diff them explicitly.\n- Do not abandon the old repository for a fresh one created from the template.\n",{"name":5,"description":7},"main","3a408474735fec0ac5d62b86fc842420a899e090",".agents\u002Fskills\u002Fblueprint-import","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia","https:\u002F\u002Fgithub.com\u002Famg0","2026-11-15 01:01:02",{"source":8,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add amg0\u002Fha_ksenia --skill blueprint-import",{"tool":35,"command":36},"bunx","bunx skills add amg0\u002Fha_ksenia --skill blueprint-import",{"tool":38,"command":39},"pnpm","pnpm dlx skills add amg0\u002Fha_ksenia --skill blueprint-import","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fblueprint-import\u002FSKILL.md",true,[43,49],{"path":44,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":47,"sourceHtmlUrl":48},"references\u002Fgit-strategy.md","REFERENCE","参考文档",3111,"https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fblueprint-import\u002Freferences\u002Fgit-strategy.md",{"path":50,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Flegacy-patterns.md",4018,"https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fblueprint-import\u002Freferences\u002Flegacy-patterns.md",[54,61,68,75,82,89],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":18},"01M7AB5RYA92GRR64CP6VMWV6Z","issues","Triage\u002Ffix the GitHub issue backlog via branch + PR, with auto-merge on green CI for this session.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.github\u002Fskills\u002Fissues\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":18},"01M7AB5G6DR1KGAPRT7JHDCD2Y","ha-translations","Add or fix user-facing strings for this Home Assistant custom integration — translations\u002Fen.json and other languages, icons.json, entity names and states, config flow labels and errors, service action descriptions, exception messages, and repair issue text. Use when asked to \"update translations\", \"add a translation key\", \"translate\", \"fix the entity name\", \"add strings for the new sensor\u002Faction\u002Foption\", \"add icons\", or when hassfest reports a missing or unused translation key. Covers the key structure, which keys each feature needs, placeholders, the icons.json contract, and the project's strict rule about only touching en.json. SYMPTOMS — load this if you are about to: write user-facing English into Python; edit a language file other than `en.json`; add an entity without its `entity.\u003Cplatform>.\u003Ckey>.name` key; skip `data_description` on a config flow field; or set `icon=` instead of using `icons.json`.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fha-translations\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M7AB56JBRMF19W4N2NG9Y5GW","ha-testing","Write, fix, or extend automated tests for this Home Assistant custom integration using pytest and pytest-homeassistant-custom-component. Use when asked to \"add tests\", \"write a test\", \"the tests fail\", \"improve coverage\", \"add a regression test\", \"test the config flow\", \"update snapshots\", or when a behavioural change needs verification before it can be called done. Covers the tests\u002F layout, the conftest.py fixtures this project needs, mocking the API client, config flow and entity test patterns, registry assertions, syrupy snapshots, freezing time, and the script\u002Ftest commands. SYMPTOMS — load this if you are about to: instantiate an entity class directly instead of asserting on `hass.states`; write a test without `enable_custom_integrations`; use `time.sleep` instead of `async_fire_time_changed`; silence a DeprecationWarning to make the suite pass; or commit a snapshot you have not read.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fha-testing\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":18},"01M7AB4V9XKDYRDCRHEA68GRZ5","ha-service-action","Add, change, or remove a Home Assistant service action (also called a \"service\" or \"action\") in this custom integration. Use when asked to \"add a service\", \"add an action\", \"expose X as a service call\", \"add a service that returns data\", \"make this callable from an automation\", or when editing services.yaml, the service_actions\u002F package, service schemas, service response data, or the async_setup() registration. Covers the services.yaml contract, probatio schemas, selectors, entity\u002Fdevice targeting, response data, translated error messages, and the Bronze quality-scale rule that actions must be registered in async_setup(). SYMPTOMS — load this if you are about to: register an action in `async_setup_entry()`; add a `services.yaml` field without a selector; add an `entity_id` field where `target:` belongs; raise a plain English `HomeAssistantError`; or swallow an action failure into a log line.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fha-service-action\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M7AB4J65REZ9T9VS9MQH73CC","ha-release","Cut a release of this Home Assistant custom integration, or prepare the commits that feed it. Use when asked to \"release\", \"cut a version\", \"bump the version\", \"prepare release notes\", \"what will the next version be\", \"write a commit message\", \"why didn't release-please pick this up\", or when reviewing whether a change is user-facing. Covers the Conventional Commits rules commitlint enforces here, how release-please derives the version and changelog, the pre-1.0 bump rules, script\u002Fversion and script\u002Frelease-notes, skipping a commit from the notes, and the commit rules this project enforces for agents. SYMPTOMS — load this if you are about to: commit without an explicit request; offer to `git push`; label a user-facing fix as `chore` (chore, like every hidden type, never reaches the changelog and never triggers a release); write a commit subject over 100 characters or in sentence case; or edit the version in `manifest.json` by hand.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fha-release\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":18},"01M7AB48WMPGDK2REBC3WS8T32","ha-quality-review","Perform a structured quality review of this Home Assistant custom integration, either as a full audit or scoped to a diff. Use when asked to \"review the integration\", \"audit the code\", \"check quality scale\", \"are we bronze\u002Fsilver\u002Fgold\", \"is this ready to release\", \"review my changes\", \"what's missing before publishing\", or before opening a pull request that touches integration code. Covers running the automated gates first, then auditing architecture, the Integration Quality Scale rules, error handling, security, performance, user experience, and documentation — and reporting findings ranked by severity with concrete fixes. SYMPTOMS — load this if you are about to: report a review that only restates linter output; claim real-device testing that did not happen; give a finding without a file path and a fix; or judge quality without running `script\u002Fhassfest` and the test suite first.","https:\u002F\u002Fgithub.com\u002Famg0\u002Fha_ksenia\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fha-quality-review\u002FSKILL.md",[],[]]