[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6MFN82EXF24T4G3PARY872H":3,"skill-same:01M6MFN82EXF24T4G3PARY872H":119},{"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":40,"linkOnly":41,"markdownContent":42,"markdownBody":43,"frontMatter":44,"refName":45,"commitSha":46,"skillDirectory":47,"forksCount":48,"repositoryHtmlUrl":49,"repositoryHomepageUrl":50,"ownerHtmlUrl":51,"ownerBlogUrl":52,"ownerDescription":53,"updatedAt":54,"install":55,"resources":68},"01M6MFN82EXF24T4G3PARY872H","codex-quota-optimizer",null,"Optimize Codex Plus\u002FPro usage before and during coding tasks. Use when the user wants to save Codex quota\u002Fcredits, reduce token or context waste, choose a cheaper model\u002Freasoning level, avoid unnecessary repo scans\u002Ftests\u002Fsubagents, or plan an implementation with the smallest reliable execution footprint. Do not use when the user explicitly prioritizes maximum quality regardless of usage.","hashgraph-online\u002Fawesome-codex-plugins","hashgraph-online","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160371962?v=4","01M6CNB6NRJVNNGN14TR01Q6D3","HOL (Hashgraph Online)","01M6JT1HSKKG9V68KQAQGBA7XP",1571,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39],"codex quota optimizer","optimizer","codex quota","codex","quota","credits","when","reduce","usage","token","optimize","do","subagents","context","repo","choose","coding","model","plan","save","tasks","2026-11-10 10:12:30",false,"---\nname: codex-quota-optimizer\ndescription: Optimize Codex Plus\u002FPro usage before and during coding tasks. Use when the user wants to save Codex quota\u002Fcredits, reduce token or context waste, choose a cheaper model\u002Freasoning level, avoid unnecessary repo scans\u002Ftests\u002Fsubagents, or plan an implementation with the smallest reliable execution footprint. Do not use when the user explicitly prioritizes maximum quality regardless of usage.\n---\n\n# Codex Quota Optimizer\n\nOptimize for **useful work per unit of Codex allowance**, not for the shortest answer.\n\n## 0. Operating principle\n\nUse the cheapest execution path that can still meet the acceptance criteria. Escalate only when evidence shows the cheaper path is insufficient.\n\nNever claim an exact remaining quota unless the current Codex client\u002Faccount exposes it. If quota state matters, tell the user to check `\u002Fstatus` or the usage dashboard, then continue with the best available strategy.\n\n### Zero-friction invariant\n\nThe optimizer must not become the overhead.\n\n- Do not make an additional model call just to classify or budget a task.\n- Do not add network requests for CQO classification, journaling, or auditing.\n- Do not put a blocking budget gate in front of normal Codex execution.\n- Do not ask the user for confirmation merely because a soft budget was exceeded.\n- Do not spawn subagents on CQO's behalf.\n- The Skill must remain fully useful when the optional `cqo` CLI is never installed or run.\n\nSee `references\u002Fzero-friction.md`.\n\n## 1. Read only what you need\n\nBefore broad exploration:\n\n1. Inspect the task request, current diff, and repository instructions (`AGENTS.md`).\n2. Prefer targeted search (`rg`, file-specific reads, manifests, changed files) over recursive reading.\n3. If the repository is unfamiliar, run `scripts\u002Frepo_snapshot.py --compact` once before opening many files.\n4. Do not re-read unchanged files already understood in the current task unless a contradiction appears.\n5. Read generated\u002Fvendor\u002Fbuild directories only when directly relevant.\n\nSee `references\u002Fcontext-policy.md` for detailed rules.\n\n## 2. Classify the task before execution\n\nAssign one class internally as part of the existing reasoning turn. Do not call another model solely to perform classification.\n\n- **XS** — one-file, mechanical, obvious acceptance criteria.\n- **S** — narrow feature\u002Ffix, 1–3 files, known pattern.\n- **M** — several files, modest design judgment or integration.\n- **L** — cross-cutting change, unclear architecture, migration, complex debugging.\n- **XL** — high-risk redesign, multi-system work, security-critical or ambiguous end-to-end task.\n\nThen choose the lowest adequate model\u002Freasoning path:\n\n- **XS\u002FS:** prefer Luna or the fastest\u002Fcheapest available setting, low reasoning.\n- **M:** prefer Terra, low\u002Fmedium reasoning.\n- **L:** Terra medium first; escalate to Sol only for unresolved design\u002Fdebugging complexity.\n- **XL:** Sol medium\u002Fhigh; Astra only when the task truly needs strongest multi-step\u002Ftool reasoning and the account exposes it.\n\nAvailability varies by plan\u002Fclient. If a named model is unavailable, choose the nearest cheaper\u002Ffaster model with the same role.\n\nDo not use Max\u002FUltra by default. Do not spawn subagents merely because they are available.\n\nSee `references\u002Fmodel-routing.md`.\n\n## 3. Use an execution budget\n\nFor every non-trivial task, maintain this internal budget:\n\n- **Discovery budget:** enough reads\u002Fsearches to locate the change surface.\n- **Edit budget:** minimum files required to satisfy acceptance criteria.\n- **Verification budget:** targeted checks first; broad checks only at the final gate or when evidence demands them.\n\nThese are **soft budgets**, never execution gates:\n\n- Stop discovery once the relevant dependency path is understood.\n- Stop editing once acceptance criteria are met; avoid opportunistic refactors.\n- Stop verification after relevant tests\u002Ftype\u002Flint checks pass unless the change is high risk.\n- If another file, test, or reasoning step is required for correctness, continue without interrupting the user.\n- If scope expands, reclassify internally and keep moving; do not request permission just because the initial budget was too small.\n\n## 4. Plan economically\n\nFor S\u002FM\u002FL\u002FXL tasks, state or maintain a compact plan with:\n\n1. target behavior,\n2. likely files,\n3. minimal verification,\n4. escalation trigger.\n\nDo not create a long speculative plan before reading the minimum repository evidence.\n\nPrefer one coherent implementation turn over repeated micro-turns that each reload context.\n\n## 5. Edit minimally\n\n- Change the smallest coherent surface.\n- Reuse existing helpers, patterns, components, and tests.\n- Avoid formatting unrelated files.\n- Avoid dependency upgrades unless required.\n- Avoid generated-file churn unless generation is part of the task.\n- Batch logically related edits rather than repeatedly opening\u002Fediting the same file.\n\nWhen a task can be solved by configuration or a small local patch, do not introduce a framework or abstraction layer.\n\n## 6. Verify in layers\n\nRun checks in this order:\n\n1. syntax\u002Fformat check for touched files,\n2. focused unit\u002Fcomponent test for touched behavior,\n3. focused type\u002Flint check,\n4. broader package\u002Fapp test only if needed,\n5. full suite only for cross-cutting\u002Fhigh-risk changes or before a release gate.\n\nDo not repeatedly run the same expensive full suite after tiny edits. If a broad test fails for an unrelated pre-existing reason, record it and continue with focused verification.\n\nSee `references\u002Ftest-policy.md`.\n\n## 7. Control context growth\n\nAt natural checkpoints, compress working state into a short handoff note:\n\n- goal,\n- decisions,\n- files changed,\n- tests run\u002Fresults,\n- remaining issue.\n\nUse the checkpoint instead of restating the entire conversation or repository history.\n\nFor long tasks, create\u002Fupdate a small local scratch note only if it prevents repeated exploration. Remove it before finalizing unless the project wants it committed.\n\n## 8. Avoid expensive anti-patterns\n\nNever do these by default:\n\n- scan the entire repository when a targeted search can locate the feature,\n- open large lockfiles, minified bundles, generated output, or vendor trees without need,\n- run full test suites before identifying the change surface,\n- use high\u002Fxhigh\u002Fmax\u002Fultra reasoning for routine edits,\n- use multiple subagents for a task one agent can do cheaply,\n- ask Codex to rewrite whole files when a patch is sufficient,\n- repeatedly summarize information already present in the current context,\n- refactor adjacent code “while here” without user value,\n- browse the web when repository\u002Flocal docs are enough,\n- perform duplicate searches with equivalent queries.\n\nSee `references\u002Fanti-patterns.md`.\n\n## 9. Escalate deliberately\n\nEscalate model\u002Freasoning\u002Ftool breadth only when at least one trigger is present:\n\n- two focused attempts fail for reasons not explained by local evidence,\n- architecture is ambiguous and wrong choice would cause broad rework,\n- failure spans multiple subsystems,\n- security\u002Fdata-loss\u002Fmigration risk is meaningful,\n- acceptance criteria conflict,\n- targeted verification exposes hidden coupling.\n\nBefore escalation, summarize what is known so the stronger model does not need to rediscover it.\n\n## 10. Plus vs Pro behavior\n\nDo not hardcode plan quotas into decisions. Use the same efficiency policy for both plans, with different default aggressiveness:\n\n- **Plus:** default to `economy` behavior — Luna\u002FTerra first, focused checks, strict scope control.\n- **Pro:** default to `balanced` behavior — Terra first, Sol when complexity justifies it; still avoid waste because Pro allowance is finite and may be shared with other agentic features.\n\nIf the user says quota is nearly exhausted, enter **emergency mode**:\n\n- no broad exploration,\n- no optional refactors,\n- no subagents,\n- cheapest adequate model,\n- one implementation path,\n- focused verification only,\n- produce a concise continuation note if completion is impossible in the current turn.\n\n## 11. User-facing output\n\nWhen useful, end with a compact **Usage choices** note containing only actionable items such as:\n\n- “This is an S task; Luna\u002Flow is sufficient.”\n- “Use Terra\u002Fmedium for implementation; reserve Sol for the migration edge case.”\n- “Run the focused test now; defer the full suite until final review.”\n\nDo not clutter every answer with quota commentary if the skill is operating successfully in the background.\n\n## 12. Optional local observability\n\nThe `cqo` CLI is optional. It provides task-level classification, soft budgets, a local journal, and a local audit:\n\n- `cqo start \u003Ctask>`\n- `cqo status`\n- `cqo audit`\n- `cqo history`\n\nThe journal is local-only under `~\u002F.cqo` (or `CQO_HOME`). It does not inspect private account pages or estimate hidden quota.\n\n**Never run `cqo` automatically just to collect analytics.** The CLI is an inspection layer, not a runtime dependency. If the user did not opt into a CQO session, normal Skill behavior continues with zero CLI overhead.\n\n## 13. Optional helper scripts\n\n- `scripts\u002Frepo_snapshot.py --compact` — compact project map without reading the whole repo.\n- `scripts\u002Fchange_scope.py` — summarize current Git change surface and suggest verification scope.\n- `scripts\u002Fcqo.py` — optional zero-network local CLI for task-level budgeting and auditing.\n\nUse scripts only when they save model context\u002Ftool calls or the user explicitly wants observability; do not run them ritualistically.\n","# Codex Quota Optimizer\n\nOptimize for **useful work per unit of Codex allowance**, not for the shortest answer.\n\n## 0. Operating principle\n\nUse the cheapest execution path that can still meet the acceptance criteria. Escalate only when evidence shows the cheaper path is insufficient.\n\nNever claim an exact remaining quota unless the current Codex client\u002Faccount exposes it. If quota state matters, tell the user to check `\u002Fstatus` or the usage dashboard, then continue with the best available strategy.\n\n### Zero-friction invariant\n\nThe optimizer must not become the overhead.\n\n- Do not make an additional model call just to classify or budget a task.\n- Do not add network requests for CQO classification, journaling, or auditing.\n- Do not put a blocking budget gate in front of normal Codex execution.\n- Do not ask the user for confirmation merely because a soft budget was exceeded.\n- Do not spawn subagents on CQO's behalf.\n- The Skill must remain fully useful when the optional `cqo` CLI is never installed or run.\n\nSee `references\u002Fzero-friction.md`.\n\n## 1. Read only what you need\n\nBefore broad exploration:\n\n1. Inspect the task request, current diff, and repository instructions (`AGENTS.md`).\n2. Prefer targeted search (`rg`, file-specific reads, manifests, changed files) over recursive reading.\n3. If the repository is unfamiliar, run `scripts\u002Frepo_snapshot.py --compact` once before opening many files.\n4. Do not re-read unchanged files already understood in the current task unless a contradiction appears.\n5. Read generated\u002Fvendor\u002Fbuild directories only when directly relevant.\n\nSee `references\u002Fcontext-policy.md` for detailed rules.\n\n## 2. Classify the task before execution\n\nAssign one class internally as part of the existing reasoning turn. Do not call another model solely to perform classification.\n\n- **XS** — one-file, mechanical, obvious acceptance criteria.\n- **S** — narrow feature\u002Ffix, 1–3 files, known pattern.\n- **M** — several files, modest design judgment or integration.\n- **L** — cross-cutting change, unclear architecture, migration, complex debugging.\n- **XL** — high-risk redesign, multi-system work, security-critical or ambiguous end-to-end task.\n\nThen choose the lowest adequate model\u002Freasoning path:\n\n- **XS\u002FS:** prefer Luna or the fastest\u002Fcheapest available setting, low reasoning.\n- **M:** prefer Terra, low\u002Fmedium reasoning.\n- **L:** Terra medium first; escalate to Sol only for unresolved design\u002Fdebugging complexity.\n- **XL:** Sol medium\u002Fhigh; Astra only when the task truly needs strongest multi-step\u002Ftool reasoning and the account exposes it.\n\nAvailability varies by plan\u002Fclient. If a named model is unavailable, choose the nearest cheaper\u002Ffaster model with the same role.\n\nDo not use Max\u002FUltra by default. Do not spawn subagents merely because they are available.\n\nSee `references\u002Fmodel-routing.md`.\n\n## 3. Use an execution budget\n\nFor every non-trivial task, maintain this internal budget:\n\n- **Discovery budget:** enough reads\u002Fsearches to locate the change surface.\n- **Edit budget:** minimum files required to satisfy acceptance criteria.\n- **Verification budget:** targeted checks first; broad checks only at the final gate or when evidence demands them.\n\nThese are **soft budgets**, never execution gates:\n\n- Stop discovery once the relevant dependency path is understood.\n- Stop editing once acceptance criteria are met; avoid opportunistic refactors.\n- Stop verification after relevant tests\u002Ftype\u002Flint checks pass unless the change is high risk.\n- If another file, test, or reasoning step is required for correctness, continue without interrupting the user.\n- If scope expands, reclassify internally and keep moving; do not request permission just because the initial budget was too small.\n\n## 4. Plan economically\n\nFor S\u002FM\u002FL\u002FXL tasks, state or maintain a compact plan with:\n\n1. target behavior,\n2. likely files,\n3. minimal verification,\n4. escalation trigger.\n\nDo not create a long speculative plan before reading the minimum repository evidence.\n\nPrefer one coherent implementation turn over repeated micro-turns that each reload context.\n\n## 5. Edit minimally\n\n- Change the smallest coherent surface.\n- Reuse existing helpers, patterns, components, and tests.\n- Avoid formatting unrelated files.\n- Avoid dependency upgrades unless required.\n- Avoid generated-file churn unless generation is part of the task.\n- Batch logically related edits rather than repeatedly opening\u002Fediting the same file.\n\nWhen a task can be solved by configuration or a small local patch, do not introduce a framework or abstraction layer.\n\n## 6. Verify in layers\n\nRun checks in this order:\n\n1. syntax\u002Fformat check for touched files,\n2. focused unit\u002Fcomponent test for touched behavior,\n3. focused type\u002Flint check,\n4. broader package\u002Fapp test only if needed,\n5. full suite only for cross-cutting\u002Fhigh-risk changes or before a release gate.\n\nDo not repeatedly run the same expensive full suite after tiny edits. If a broad test fails for an unrelated pre-existing reason, record it and continue with focused verification.\n\nSee `references\u002Ftest-policy.md`.\n\n## 7. Control context growth\n\nAt natural checkpoints, compress working state into a short handoff note:\n\n- goal,\n- decisions,\n- files changed,\n- tests run\u002Fresults,\n- remaining issue.\n\nUse the checkpoint instead of restating the entire conversation or repository history.\n\nFor long tasks, create\u002Fupdate a small local scratch note only if it prevents repeated exploration. Remove it before finalizing unless the project wants it committed.\n\n## 8. Avoid expensive anti-patterns\n\nNever do these by default:\n\n- scan the entire repository when a targeted search can locate the feature,\n- open large lockfiles, minified bundles, generated output, or vendor trees without need,\n- run full test suites before identifying the change surface,\n- use high\u002Fxhigh\u002Fmax\u002Fultra reasoning for routine edits,\n- use multiple subagents for a task one agent can do cheaply,\n- ask Codex to rewrite whole files when a patch is sufficient,\n- repeatedly summarize information already present in the current context,\n- refactor adjacent code “while here” without user value,\n- browse the web when repository\u002Flocal docs are enough,\n- perform duplicate searches with equivalent queries.\n\nSee `references\u002Fanti-patterns.md`.\n\n## 9. Escalate deliberately\n\nEscalate model\u002Freasoning\u002Ftool breadth only when at least one trigger is present:\n\n- two focused attempts fail for reasons not explained by local evidence,\n- architecture is ambiguous and wrong choice would cause broad rework,\n- failure spans multiple subsystems,\n- security\u002Fdata-loss\u002Fmigration risk is meaningful,\n- acceptance criteria conflict,\n- targeted verification exposes hidden coupling.\n\nBefore escalation, summarize what is known so the stronger model does not need to rediscover it.\n\n## 10. Plus vs Pro behavior\n\nDo not hardcode plan quotas into decisions. Use the same efficiency policy for both plans, with different default aggressiveness:\n\n- **Plus:** default to `economy` behavior — Luna\u002FTerra first, focused checks, strict scope control.\n- **Pro:** default to `balanced` behavior — Terra first, Sol when complexity justifies it; still avoid waste because Pro allowance is finite and may be shared with other agentic features.\n\nIf the user says quota is nearly exhausted, enter **emergency mode**:\n\n- no broad exploration,\n- no optional refactors,\n- no subagents,\n- cheapest adequate model,\n- one implementation path,\n- focused verification only,\n- produce a concise continuation note if completion is impossible in the current turn.\n\n## 11. User-facing output\n\nWhen useful, end with a compact **Usage choices** note containing only actionable items such as:\n\n- “This is an S task; Luna\u002Flow is sufficient.”\n- “Use Terra\u002Fmedium for implementation; reserve Sol for the migration edge case.”\n- “Run the focused test now; defer the full suite until final review.”\n\nDo not clutter every answer with quota commentary if the skill is operating successfully in the background.\n\n## 12. Optional local observability\n\nThe `cqo` CLI is optional. It provides task-level classification, soft budgets, a local journal, and a local audit:\n\n- `cqo start \u003Ctask>`\n- `cqo status`\n- `cqo audit`\n- `cqo history`\n\nThe journal is local-only under `~\u002F.cqo` (or `CQO_HOME`). It does not inspect private account pages or estimate hidden quota.\n\n**Never run `cqo` automatically just to collect analytics.** The CLI is an inspection layer, not a runtime dependency. If the user did not opt into a CQO session, normal Skill behavior continues with zero CLI overhead.\n\n## 13. Optional helper scripts\n\n- `scripts\u002Frepo_snapshot.py --compact` — compact project map without reading the whole repo.\n- `scripts\u002Fchange_scope.py` — summarize current Git change surface and suggest verification scope.\n- `scripts\u002Fcqo.py` — optional zero-network local CLI for task-level budgeting and auditing.\n\nUse scripts only when they save model context\u002Ftool calls or the user explicitly wants observability; do not run them ritualistically.\n",{"name":5,"description":7},"main","2b63f856ce19c2a797929cacabd9e0fc936812fa","plugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer",490,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins","https:\u002F\u002Fhol.org\u002Fplugins\u002Fbest-codex-plugins","https:\u002F\u002Fgithub.com\u002Fhashgraph-online","https:\u002F\u002Fhol.org","Open-source standards and developer tools for interoperable AI agents: discovery, identity, communication, payments, privacy and security.","2026-11-14 02:12:02",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add hashgraph-online\u002Fawesome-codex-plugins --skill codex-quota-optimizer",{"tool":61,"command":62},"bunx","bunx skills add hashgraph-online\u002Fawesome-codex-plugins --skill codex-quota-optimizer",{"tool":64,"command":65},"pnpm","pnpm dlx skills add hashgraph-online\u002Fawesome-codex-plugins --skill codex-quota-optimizer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002FSKILL.md",true,[69,75,81,85,89,93,97,101,105,111,115],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"agents\u002Fopenai.yaml","REFERENCE","参考文档",521,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Fagents\u002Fopenai.yaml",{"path":76,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":79,"sourceHtmlUrl":80},"assets\u002Ficon.svg","ASSET","资产文件",571,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Fassets\u002Ficon.svg",{"path":82,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":83,"sourceHtmlUrl":84},"references\u002Fanti-patterns.md",838,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Fanti-patterns.md",{"path":86,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":87,"sourceHtmlUrl":88},"references\u002Fcontext-policy.md",971,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Fcontext-policy.md",{"path":90,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":91,"sourceHtmlUrl":92},"references\u002Fcurrent-openai-notes.md",1241,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Fcurrent-openai-notes.md",{"path":94,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":95,"sourceHtmlUrl":96},"references\u002Fmodel-routing.md",1155,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Fmodel-routing.md",{"path":98,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":99,"sourceHtmlUrl":100},"references\u002Ftest-policy.md",815,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Ftest-policy.md",{"path":102,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":103,"sourceHtmlUrl":104},"references\u002Fzero-friction.md",2640,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Freferences\u002Fzero-friction.md",{"path":106,"resourceType":107,"resourceTypeDesc":108,"sizeBytes":109,"sourceHtmlUrl":110},"scripts\u002Fchange_scope.py","SCRIPT","脚本",1629,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Fscripts\u002Fchange_scope.py",{"path":112,"resourceType":107,"resourceTypeDesc":108,"sizeBytes":113,"sourceHtmlUrl":114},"scripts\u002Fcqo.py",18306,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Fscripts\u002Fcqo.py",{"path":116,"resourceType":107,"resourceTypeDesc":108,"sizeBytes":117,"sourceHtmlUrl":118},"scripts\u002Frepo_snapshot.py",3217,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fctdaniel\u002Fcodex-quota-optimizer\u002Fskills\u002Fcodex-quota-optimizer\u002Fscripts\u002Frepo_snapshot.py",[120,128,135,142,168,193],{"skillNo":121,"name":122,"nameZh":6,"description":123,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":124,"categories":125,"keywords":126,"publishedAt":127},"01M746ZWACVKMZ1BJXJZHYJSSX","migrate-to-e2e","Moves an existing Playwright, Cypress, Selenium, Appium, Detox, Maestro, or TesterArmy suite to e2e one slice at a time, with the old suite running beside the new one until each ported test is proven. Keeps a MIGRATION.md progress file so any later session resumes where the last stopped. Optionally plants a defect to show both suites catch it. Use for \u002Fmigrate-to-e2e, \"migrate our tests to e2e\", \"port these Maestro flows\", \"move our TesterArmy tests\", or when a project has an e2e config and an older suite still running.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Ftester-army\u002Fe2e\u002Fskills\u002Fmigrate-to-e2e\u002FSKILL.md",[],[],"2026-11-11 15:11:14",{"skillNo":129,"name":130,"nameZh":6,"description":131,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":132,"categories":133,"keywords":134,"publishedAt":127},"01M746ZPTAX1Q0FBKAB1CYND7E","e2e","Writes, runs, and debugs end-to-end tests for web and mobile apps with e2e, the agentic test runner, which mixes agent.act goals with exact locator assertions on a Playwright browser or an iOS or Android device. Also drives a project's running app live over the e2e MCP server to look at a screen, check a UI change, take a screenshot or record a video, and try a locator; and runs bug bashes, parallel e2e explore runs with each bug proven by a repro test. Use when a project depends on e2e or has an e2e.config.ts, when asked for end-to-end, browser, mobile, or agentic UI tests, to check a change in the running app, to QA or bug bash an app or branch, or when an e2e run fails.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Ftester-army\u002Fe2e\u002Fskills\u002Fe2e\u002FSKILL.md",[],[],{"skillNo":136,"name":137,"nameZh":6,"description":138,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":139,"categories":140,"keywords":141,"publishedAt":127},"01M746ZJN784HVHMDP9MQGCSGG","create-verification-skill","Generate a project-local verify-\u003Capp> skill on top of e2e, so any coding agent can launch the app, check it is worth driving, exercise a feature the way a user does, keep evidence, and bug bash it. Writes the skill, a feature map with one explore charter per feature, and proves it once. Use for \u002Fcreate-verification-skill, \"make a verification skill for this repo\", or when a project has no scripted way to prove UI behavior.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Ftester-army\u002Fe2e\u002Fskills\u002Fcreate-verification-skill\u002FSKILL.md",[],[],{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":146,"categories":147,"keywords":148,"publishedAt":40},"01M716A6ETNB9B1SQ3GXZC75VR","deja-search","deja-vu memory — search the user's past AI coding sessions with the deja CLI. Holds what they already settled and never wrote into the repo (naming rules, banned or already-tried tools, required checks before merge, preferred values, who owns what). Use before your first edit in a task, before you call a change done or ready to merge, before you recommend a tool, library, design or reviewer, before re-debugging an error that may already be solved, when they say 'didn't we fix this before', 'what did we decide about X' or 'I already have that', and before telling them something on this machine does not exist.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fvshulcz\u002Fdeja-vu\u002Fskills\u002Fdeja-search\u002FSKILL.md",[],[149,150,151,152,153,154,155,156,157,158,159,25,160,161,162,163,164,165,166,167],"search","deja search","never","error","ready","memory","re","decide","fix","design","debugging","you","done","reviewer","io","change","cli","merge","past",{"skillNo":169,"name":170,"nameZh":6,"description":171,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":172,"categories":173,"keywords":174,"publishedAt":40},"01M716A3J2W5SENSE9VGMXB375","deja-reflect","End-of-session reflection and memory curation with deja-vu. Review what was decided, what commands broke and what repaired them, and record durable project rules or decisions with deja remember so the next session starts with them.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fvshulcz\u002Fdeja-vu\u002Fskills\u002Fdeja-reflect\u002FSKILL.md",[],[175,176,177,163,178,179,180,181,182,183,184,185,186,187,154,188,189,190,191,192],"deja reflect","reflect","review","with","record","github","next","https","node","Rules","decisions","brew","remember","project rules","project","session","end","reflection",{"skillNo":194,"name":195,"nameZh":6,"description":196,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":197,"categories":198,"keywords":210,"publishedAt":40},"01M715T3RR4BAYSR57HPPTJ0PF","revid","Make, edit, export and post AI videos with Revid (revid.ai) through the Revid MCP server. Use whenever the user wants a short-form or long-form video made (from an idea, a script, a URL, an article, a PDF, a tweet, a product page, their own clips, or a song), wants to find what is going viral on TikTok, YouTube, Instagram or X before making something, wants to edit or re-export an existing Revid project, generate a standalone AI image or clip, make a series episode, or publish and schedule a finished video to connected social accounts. Covers setup (hosted MCP with OAuth, or an API key), the 36 tools, credits, polling and the rules for publishing. Routes to the more specific revid-* skills when one fits.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Ftmakersas\u002Frevid-skills\u002Fskills\u002Frevid\u002FSKILL.md",[199,204,207],{"categoryNo":200,"categoryName":201,"taxonomyType":202,"taxonomyTypeDesc":203},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",{"categoryNo":205,"categoryName":206,"taxonomyType":202,"taxonomyTypeDesc":203},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":208,"categoryName":209,"taxonomyType":202,"taxonomyTypeDesc":203},"01JBSEEDCAT000000000000010","后端与接口",[195,211,212,213,214,215,216,189,217,155,218,219,220,221,25,222,223,224,225,226],"URL","Video","short-form","standalone","PDF","viral","tiktok","short","social","MCP","song","image","mcp server","schedule","api","publish"]