[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7E84BT64PS1W1PFVJS8G3VQ":3,"skill-same:01M7E84BT64PS1W1PFVJS8G3VQ":76},{"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":34,"commitSha":35,"skillDirectory":36,"forksCount":14,"repositoryHtmlUrl":37,"repositoryHomepageUrl":6,"ownerHtmlUrl":38,"ownerBlogUrl":39,"ownerDescription":40,"updatedAt":41,"install":42,"resources":55},"01M7E84BT64PS1W1PFVJS8G3VQ","terraform-bumping-terraform",null,"Use when Terraform or OpenTofu providers, CLI pins, or lockfiles are outdated or behind; bump or upgrade them across root modules.","tony\u002Fskills","tony","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F26336?v=4","01M72E5RNJ7SQV8F9KBM5RKRMN","Tony Narlock","01M7E7MBN5C6YZ4JJJ5Y82KBJG",0,"MIT","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: terraform-bumping-terraform\ndescription: >-\n  Use when Terraform or OpenTofu providers, CLI pins, or lockfiles are\n  outdated or behind; bump or upgrade them across root modules.\nallowed-tools: [\"Bash\", \"Read\", \"Grep\", \"Glob\", \"Edit\", \"Write\", \"AskUserQuestion\"]\nmetadata:\n  source: \"plugins\u002Fterraform\u002Fskills\u002Fbumping-terraform\u002FSKILL.md\"\n---\n\n# Bumping Terraform\n\nFind the root modules, move every declaration of a version together,\nrefresh each root module's lock file without narrowing it, and verify\nthe selected version changed. This skill classifies the request and hands\nit to one of three skills; the conventions below apply to whichever runs.\n\n## Conversational requests\n\nThe three slash-command skills are gated, so the model cannot invoke them.\nMatch the request to one, then read its file and follow it from its\nProcedure through its Output, taking its arguments from the request's\nwording:\n\n- A provider to move: the `terraform-bump-provider` skill.\n- The CLI version to move: the `terraform-bump-terraform` skill.\n- Locks to refresh inside existing constraints:\n  the `terraform-refresh-lock` skill.\n\nA request that names none (\"everything here is behind\") is a provider\nbump with no provider named unless the user says the CLI: list the\ndeclared providers and ask which (headless default: stop). One run does\none of them; offer the others as next steps.\n\n## Conventions for every run\n\n- Name the current branch in the plan. Before the run's first edit, run\n  `git status --porcelain` and print the result: that list is the\n  changes the run did not make. Never rely on it from memory in later\n  commands. Commit only this run's paths with `git commit -- \u003Cpaths>`,\n  after `git add \u003Cpath>` for any new file such as a lock file (`git\n  commit -- \u003Cuntracked path>` fails). Stop when a path to commit is\n  already in the recorded list. Also before the first edit, record each\n  gate's result on the original tree, as [Verification\n  gates](references\u002Flock-and-init.md#verification-gates) describes.\n- Script paths: `references\u002Flockctl.py` resolves from a skill's\n  directory (a reference writes `..\u002Fscripts\u002F`). Print the absolute path\n  once and write it literally in each command, since shell variables do\n  not persist between Bash calls.\n- Run `init`, `providers lock`, `lockctl.py diff`, each repository gate,\n  and `git commit` in separate Bash calls, one command per call. Read\n  that command's returned exit status before continuing. Read long\n  output from the host's saved result separately; a display filter or\n  trailing command can hide the operation's failure.\n- After `git commit`, read it back in separate calls: `git show\n  --name-status --format=%s HEAD` lists exactly this run's paths, and\n  `git status --porcelain` adds nothing to the recorded list beyond\n  `init`'s `.terraform\u002F` cache. Report the commit only when both hold;\n  otherwise report what differs.\n- Commit bodies say what moved and why, and cite release notes at the\n  tag the registry reported only when the run fetched that page; with\n  none, the body names no link. Take the subject format from the\n  project's AGENTS.md or CLAUDE.md. Write them per\n  `references\u002Fprose-rules.md`.\n- Write each commit message to a temp file with the file tool and commit\n  with `git commit -F \u003Cfile> -- \u003Cpaths>`. Messages quote registry text\n  (sources, tags, links), which a double-quoted `-m` or a heredoc would\n  expand or end early.\n- Commit messages carry no `Co-Authored-By`, no \"Generated with\" or other\n  tool trailers, and no emoji.\n- Where there is no interactive question tool (Codex, headless runs),\n  state each question in the reply with the narrower answer as the\n  default (fewest modules, `--no-commit`), without remarking that no\n  question tool exists. A headless run presents the\n  plan and stops, unless the request itself authorises the edit (\"go\n  ahead and bump it\"); that covers the edit only, never a commit or\n  push.\n- Reply in the shape each skill's Output declares, as plain text outside\n  any code fence, with plain bullets and no bold (never\n  `- **Label:** ...`), and no section that would hold nothing. Name paths\n  relative to the repository, never local absolute or temporary paths,\n  and name tools by name (`terraform`), never by install path. Name no\n  helper script in a reply; say what it found.\n- Where no question tool exists, close with `Next steps:` and a plain\n  bulleted list of the options, in that form every time, then stop. A\n  stopped run is the exception: it has no panel and no `Next steps:`\n  list.\n- A reply's `## Layout` names what the user can act on: the roots and\n  children found and any settled classification. Name the driving tool\n  only when both `terraform` and `tofu` are present. Leave out the branch\n  and any tool that is not installed.\n- Name no version the registry has not confirmed, and make no claim about\n  what a newer release changes (such as a major jump) before one is\n  confirmed. A version the user named that the registry cannot confirm\n  (unreachable, or absent from the release list) is a stopped run unless\n  the user said they confirmed it; the next step offers that confirmation\n  or access to the registry.\n- Output rules for every reply: state results, not the rules behind them.\n  Do not mention commit-message rules, trailers, or this skill's\n  conventions. Do not recap a default taken (\"I've taken the default\"); a\n  next-step option \"stop here\" is fine. Do not remark on the missing question tool. Do not\n  write \"X, not Y\" contrasts or filler such as \"nothing needed\n  classifying by you\". A bullet states a finding, not scope the plan\n  already gave. The hero is the line the skill's shape gives with\n  nothing appended. A completed report states gate results in\n  `## Verification`; a halted run uses the stopped form below. Every\n  reply omits what the repository lacks (\"no other file pins the CLI\", \"no child modules\n  found\", \"tflint isn't installed\", \"the repository defines no other\n  gates\") and never states what the run will do on its own later (\"I'll\n  run the audit again\"). A completed `--no-commit` report does not say\n  nothing was committed.\n- A run that cannot proceed (nothing declared, no version named and the\n  release list unreachable, a dirty tree it cannot work around, a lock\n  that failed after the edits, a gate the edits plausibly broke) ends\n  with `⚠ Stopped: \u003Cone result sentence>`, then the step the user must\n  take. The result sentence says what state the repository is in: `so\n  nothing was edited` before the first edit, or `so nothing was\n  committed` after it. A run stopped after its edits leaves them in the\n  working tree, and the step offers reverting them (`git restore --\n  \u003Cpaths>`) first.\n\n  The reply has no sections, no recap, and no numbered next-step list.\n  The step is one or two sentences and ends with its request; it states\n  nothing the run will do afterwards (\"I'll show the plan\", \"then I'll\n  rerun it\", \"so I can ...\"). A stop caused by missing configuration says\n  so once, in the result sentence, and does not list the files or commits\n  searched for.\n\n  Bullets are optional, at most three, and each states a repository fact\n  the step depends on, usually the declared pins that disagree. Write no\n  bullets when there is no such fact. Never write a bullet about:\n\n  - lock files, backends, call graphs, or the installed CLI, or whether a\n    pin or installed version satisfies the target;\n  - the branch, history, untracked files, or the driving tool;\n  - what the run would do after the edit, or anything the repository lacks\n    (\"no lock file\", \"no CI file runs providers lock\");\n  - a question or request.\n\n  Exceptions: the result sentence names roots outside `--root-module` that\n  call a child module the run would edit, and a stop after the edits may\n  carry one bullet with the new constraint and the lock's selected version\n  as they stand (\"the lock selects 3.7.0 with 1 h1 checksum, 2 before\").\n  Name no sandbox or network block; say the registry or mirror was\n  unreachable.\n\n  Reply in this shape, as plain text outside any code fence, with plain bullets, no bold (never `- **Label:** ...`), and no local paths:\n\n  ```text\n  ⚠ Stopped: I could not confirm hashicorp\u002Faws 6.57.0 because registry.terraform.io was unreachable, so nothing was edited.\n\n  Tell me you confirmed 6.57.0 yourself, or allow registry.terraform.io.\n\n  - envs\u002Fprod and modules\u002Fnetwork both declare hashicorp\u002Faws, at ~> 6.56.0 and ~> 6.55.0, so the pins disagree.\n  ```\n\n  After edits, with the lock restored and the constraint edit remaining:\n\n  ```text\n  ⚠ Stopped: infra\u002Fprod cannot resolve random 3.7.0 from its mirror, so nothing was committed.\n\n  Revert the constraint edit with `git restore -- infra\u002Fprod\u002Fversions.tf`, or supply the required provider packages.\n\n  - infra\u002Fprod declares ~> 3.7.0; its committed lock selects 3.6.0 with 2 h1 records.\n  ```\n\n- `--audit-only` writes nothing and is a result, not a stop. Its hero is\n  `Audit only: \u003Cresult>` with no `⚠`, the skill's own sections follow\n  (omit the empty ones), and no sentence restates that nothing was\n  written. In an audit `## Held back` lists only what a constraint or\n  guard is already blocking, never constraints that merely permit a\n  newer release. An audit with nothing to audit (no Terraform files) is a\n  stopped run.\n- A target the user confirmed is reported as \"confirmed by you\", with no\n  reason attached. `## Drift` states only what `## Sites` and `## Locks`\n  do not already say.\n- When a confirmed target crosses a major version, link the project's\n  upgrade guide in the report, or say none was found.\n\n## Core principle\n\nConstraints from every module in a configuration must all be satisfied at\nonce, so the unit of work is the configuration, not the file. Every\ndeclaration moves in one commit, or none does.\n\n## Scope\n\nThe current repository. Discover its root modules and confirm the set with\nthe user when there is more than one. When `--root-module` limits the run, list every root outside it that\ncalls a child module the run would edit (or declares the target pin). Such a root would be left\ninconsistent, its lock failing `init`, so stop and name them, or widen the scope on the user's yes\n(headless: stop). Never widen scope as a side effect:\nno repository-wide operator normalising, no invented platform names,\nno reformatting of modules the change did not touch.\n\n## Orchestration Plan\n\nEnter plan mode before the first edit. In Claude Code call `EnterPlanMode`; in Cursor, Codex, or Gemini use `\u002Fplan` or `Shift+Tab`. Where plan mode is unavailable, present the plan as text.\n\nPresent a plan covering:\n\n- The root modules in scope, and any directory whose classification the\n  user settled.\n- The target version, and the evidence it exists.\n- Every pin site to be edited, the operator each one keeps, and the\n  sites left alone with the reason.\n- Which lock files will be refreshed, each provider's `h1:` count, and\n  any platform names declared by the repository or confirmed by the user.\n- The commit this produces (constraints and locks together) and whether\n  it will be pushed.\n- Drift found that this run will not fix.\n\nWait for approval, then exit plan mode before the first edit. Without plan\nmode, finish discovery, resolution, and inventory, and confirm scope with\nthe user before writing anything. Headless, the plan is the reply.\n\n\n\n## Portability notes\n\n- Bundled files — every relative path in this skill points at a file shipped inside this skill directory. Read them from here, not from the host's plugin tree.\n","# Bumping Terraform\n\nFind the root modules, move every declaration of a version together,\nrefresh each root module's lock file without narrowing it, and verify\nthe selected version changed. This skill classifies the request and hands\nit to one of three skills; the conventions below apply to whichever runs.\n\n## Conversational requests\n\nThe three slash-command skills are gated, so the model cannot invoke them.\nMatch the request to one, then read its file and follow it from its\nProcedure through its Output, taking its arguments from the request's\nwording:\n\n- A provider to move: the `terraform-bump-provider` skill.\n- The CLI version to move: the `terraform-bump-terraform` skill.\n- Locks to refresh inside existing constraints:\n  the `terraform-refresh-lock` skill.\n\nA request that names none (\"everything here is behind\") is a provider\nbump with no provider named unless the user says the CLI: list the\ndeclared providers and ask which (headless default: stop). One run does\none of them; offer the others as next steps.\n\n## Conventions for every run\n\n- Name the current branch in the plan. Before the run's first edit, run\n  `git status --porcelain` and print the result: that list is the\n  changes the run did not make. Never rely on it from memory in later\n  commands. Commit only this run's paths with `git commit -- \u003Cpaths>`,\n  after `git add \u003Cpath>` for any new file such as a lock file (`git\n  commit -- \u003Cuntracked path>` fails). Stop when a path to commit is\n  already in the recorded list. Also before the first edit, record each\n  gate's result on the original tree, as [Verification\n  gates](references\u002Flock-and-init.md#verification-gates) describes.\n- Script paths: `references\u002Flockctl.py` resolves from a skill's\n  directory (a reference writes `..\u002Fscripts\u002F`). Print the absolute path\n  once and write it literally in each command, since shell variables do\n  not persist between Bash calls.\n- Run `init`, `providers lock`, `lockctl.py diff`, each repository gate,\n  and `git commit` in separate Bash calls, one command per call. Read\n  that command's returned exit status before continuing. Read long\n  output from the host's saved result separately; a display filter or\n  trailing command can hide the operation's failure.\n- After `git commit`, read it back in separate calls: `git show\n  --name-status --format=%s HEAD` lists exactly this run's paths, and\n  `git status --porcelain` adds nothing to the recorded list beyond\n  `init`'s `.terraform\u002F` cache. Report the commit only when both hold;\n  otherwise report what differs.\n- Commit bodies say what moved and why, and cite release notes at the\n  tag the registry reported only when the run fetched that page; with\n  none, the body names no link. Take the subject format from the\n  project's AGENTS.md or CLAUDE.md. Write them per\n  `references\u002Fprose-rules.md`.\n- Write each commit message to a temp file with the file tool and commit\n  with `git commit -F \u003Cfile> -- \u003Cpaths>`. Messages quote registry text\n  (sources, tags, links), which a double-quoted `-m` or a heredoc would\n  expand or end early.\n- Commit messages carry no `Co-Authored-By`, no \"Generated with\" or other\n  tool trailers, and no emoji.\n- Where there is no interactive question tool (Codex, headless runs),\n  state each question in the reply with the narrower answer as the\n  default (fewest modules, `--no-commit`), without remarking that no\n  question tool exists. A headless run presents the\n  plan and stops, unless the request itself authorises the edit (\"go\n  ahead and bump it\"); that covers the edit only, never a commit or\n  push.\n- Reply in the shape each skill's Output declares, as plain text outside\n  any code fence, with plain bullets and no bold (never\n  `- **Label:** ...`), and no section that would hold nothing. Name paths\n  relative to the repository, never local absolute or temporary paths,\n  and name tools by name (`terraform`), never by install path. Name no\n  helper script in a reply; say what it found.\n- Where no question tool exists, close with `Next steps:` and a plain\n  bulleted list of the options, in that form every time, then stop. A\n  stopped run is the exception: it has no panel and no `Next steps:`\n  list.\n- A reply's `## Layout` names what the user can act on: the roots and\n  children found and any settled classification. Name the driving tool\n  only when both `terraform` and `tofu` are present. Leave out the branch\n  and any tool that is not installed.\n- Name no version the registry has not confirmed, and make no claim about\n  what a newer release changes (such as a major jump) before one is\n  confirmed. A version the user named that the registry cannot confirm\n  (unreachable, or absent from the release list) is a stopped run unless\n  the user said they confirmed it; the next step offers that confirmation\n  or access to the registry.\n- Output rules for every reply: state results, not the rules behind them.\n  Do not mention commit-message rules, trailers, or this skill's\n  conventions. Do not recap a default taken (\"I've taken the default\"); a\n  next-step option \"stop here\" is fine. Do not remark on the missing question tool. Do not\n  write \"X, not Y\" contrasts or filler such as \"nothing needed\n  classifying by you\". A bullet states a finding, not scope the plan\n  already gave. The hero is the line the skill's shape gives with\n  nothing appended. A completed report states gate results in\n  `## Verification`; a halted run uses the stopped form below. Every\n  reply omits what the repository lacks (\"no other file pins the CLI\", \"no child modules\n  found\", \"tflint isn't installed\", \"the repository defines no other\n  gates\") and never states what the run will do on its own later (\"I'll\n  run the audit again\"). A completed `--no-commit` report does not say\n  nothing was committed.\n- A run that cannot proceed (nothing declared, no version named and the\n  release list unreachable, a dirty tree it cannot work around, a lock\n  that failed after the edits, a gate the edits plausibly broke) ends\n  with `⚠ Stopped: \u003Cone result sentence>`, then the step the user must\n  take. The result sentence says what state the repository is in: `so\n  nothing was edited` before the first edit, or `so nothing was\n  committed` after it. A run stopped after its edits leaves them in the\n  working tree, and the step offers reverting them (`git restore --\n  \u003Cpaths>`) first.\n\n  The reply has no sections, no recap, and no numbered next-step list.\n  The step is one or two sentences and ends with its request; it states\n  nothing the run will do afterwards (\"I'll show the plan\", \"then I'll\n  rerun it\", \"so I can ...\"). A stop caused by missing configuration says\n  so once, in the result sentence, and does not list the files or commits\n  searched for.\n\n  Bullets are optional, at most three, and each states a repository fact\n  the step depends on, usually the declared pins that disagree. Write no\n  bullets when there is no such fact. Never write a bullet about:\n\n  - lock files, backends, call graphs, or the installed CLI, or whether a\n    pin or installed version satisfies the target;\n  - the branch, history, untracked files, or the driving tool;\n  - what the run would do after the edit, or anything the repository lacks\n    (\"no lock file\", \"no CI file runs providers lock\");\n  - a question or request.\n\n  Exceptions: the result sentence names roots outside `--root-module` that\n  call a child module the run would edit, and a stop after the edits may\n  carry one bullet with the new constraint and the lock's selected version\n  as they stand (\"the lock selects 3.7.0 with 1 h1 checksum, 2 before\").\n  Name no sandbox or network block; say the registry or mirror was\n  unreachable.\n\n  Reply in this shape, as plain text outside any code fence, with plain bullets, no bold (never `- **Label:** ...`), and no local paths:\n\n  ```text\n  ⚠ Stopped: I could not confirm hashicorp\u002Faws 6.57.0 because registry.terraform.io was unreachable, so nothing was edited.\n\n  Tell me you confirmed 6.57.0 yourself, or allow registry.terraform.io.\n\n  - envs\u002Fprod and modules\u002Fnetwork both declare hashicorp\u002Faws, at ~> 6.56.0 and ~> 6.55.0, so the pins disagree.\n  ```\n\n  After edits, with the lock restored and the constraint edit remaining:\n\n  ```text\n  ⚠ Stopped: infra\u002Fprod cannot resolve random 3.7.0 from its mirror, so nothing was committed.\n\n  Revert the constraint edit with `git restore -- infra\u002Fprod\u002Fversions.tf`, or supply the required provider packages.\n\n  - infra\u002Fprod declares ~> 3.7.0; its committed lock selects 3.6.0 with 2 h1 records.\n  ```\n\n- `--audit-only` writes nothing and is a result, not a stop. Its hero is\n  `Audit only: \u003Cresult>` with no `⚠`, the skill's own sections follow\n  (omit the empty ones), and no sentence restates that nothing was\n  written. In an audit `## Held back` lists only what a constraint or\n  guard is already blocking, never constraints that merely permit a\n  newer release. An audit with nothing to audit (no Terraform files) is a\n  stopped run.\n- A target the user confirmed is reported as \"confirmed by you\", with no\n  reason attached. `## Drift` states only what `## Sites` and `## Locks`\n  do not already say.\n- When a confirmed target crosses a major version, link the project's\n  upgrade guide in the report, or say none was found.\n\n## Core principle\n\nConstraints from every module in a configuration must all be satisfied at\nonce, so the unit of work is the configuration, not the file. Every\ndeclaration moves in one commit, or none does.\n\n## Scope\n\nThe current repository. Discover its root modules and confirm the set with\nthe user when there is more than one. When `--root-module` limits the run, list every root outside it that\ncalls a child module the run would edit (or declares the target pin). Such a root would be left\ninconsistent, its lock failing `init`, so stop and name them, or widen the scope on the user's yes\n(headless: stop). Never widen scope as a side effect:\nno repository-wide operator normalising, no invented platform names,\nno reformatting of modules the change did not touch.\n\n## Orchestration Plan\n\nEnter plan mode before the first edit. In Claude Code call `EnterPlanMode`; in Cursor, Codex, or Gemini use `\u002Fplan` or `Shift+Tab`. Where plan mode is unavailable, present the plan as text.\n\nPresent a plan covering:\n\n- The root modules in scope, and any directory whose classification the\n  user settled.\n- The target version, and the evidence it exists.\n- Every pin site to be edited, the operator each one keeps, and the\n  sites left alone with the reason.\n- Which lock files will be refreshed, each provider's `h1:` count, and\n  any platform names declared by the repository or confirmed by the user.\n- The commit this produces (constraints and locks together) and whether\n  it will be pushed.\n- Drift found that this run will not fix.\n\nWait for approval, then exit plan mode before the first edit. Without plan\nmode, finish discovery, resolution, and inventory, and confirm scope with\nthe user before writing anything. Headless, the plan is the reply.\n\n\n\n## Portability notes\n\n- Bundled files — every relative path in this skill points at a file shipped inside this skill directory. Read them from here, not from the host's plugin tree.\n",{"name":5,"metadata":24,"description":7,"allowed-tools":26},{"source":25},"plugins\u002Fterraform\u002Fskills\u002Fbumping-terraform\u002FSKILL.md",[27,28,29,30,31,32,33],"Bash","Read","Grep","Glob","Edit","Write","AskUserQuestion","main","5fbeb3b47beedea13813be7b7023f736a8d97a65",".agents\u002Fskills\u002Fterraform-bumping-terraform","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills","https:\u002F\u002Fgithub.com\u002Ftony","https:\u002F\u002Ftony.sh","lighthouse keeper","2026-11-08 17:57:34",{"source":8,"skillName":5,"commands":43,"prompt":53,"zipAvailable":54},[44,47,50],{"tool":45,"command":46},"npx","npx skills add tony\u002Fskills --skill terraform-bumping-terraform",{"tool":48,"command":49},"bunx","bunx skills add tony\u002Fskills --skill terraform-bumping-terraform",{"tool":51,"command":52},"pnpm","pnpm dlx skills add tony\u002Fskills --skill terraform-bumping-terraform","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002FSKILL.md",true,[56,62,66,72],{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"agents\u002Fopenai.yaml","REFERENCE","参考文档",270,"https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002Fagents\u002Fopenai.yaml",{"path":63,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Flock-and-init.md",14822,"https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002Freferences\u002Flock-and-init.md",{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Flockctl.py","SCRIPT","脚本",10050,"https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002Freferences\u002Flockctl.py",{"path":73,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Fprose-rules.md",4644,"https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fterraform-bumping-terraform\u002Freferences\u002Fprose-rules.md",[77,84,91,98,105,112],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7E8NCSHXV2JPW4J61R0TQS7","serene-bliss","Use when asked to make a CLI --help, docs page, error message, or tooling UX more serene or calmer for developer experience; returns text, no edits.","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Fserene-bliss\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7E8N87MAVJD8RRQKQB1ZPS6","review","Weave code review: run independent adversarial reviews in parallel, then synthesize findings","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Freview\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7E8N402XG4ZBFNACQ413163","refine","Use when asked to refine a draft, iterate on code until cleaner, or polish a README through repeated critique and improvement.","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Frefine\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7E8N01KB86A7136P00GVKJW","prompt","Weave prompt: compare independent adversarial implementations, then adopt the single best one","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Fprompt\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":19},"01M7E8MW7YJW2PCMDW3G9RTGNK","plan","Weave planning: compare independent adversarial plans, then synthesize the best one","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Fplan\u002FSKILL.md",[],[],{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"publishedAt":19},"01M7E8MSBYXNCV4ZYWDS4PFJXK","fix-review","Fix weave review findings: validate, add test coverage, fix, and commit each as atomic changes","https:\u002F\u002Fgithub.com\u002Ftony\u002Fskills\u002Fblob\u002Fmain\u002Fplugins\u002Fweave\u002Fskills\u002Ffix-review\u002FSKILL.md",[],[]]