[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6WZZZHCTPZ8YF9564J3QJVT":3,"skill-same:01M6WZZZHCTPZ8YF9564J3QJVT":95},{"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":31,"publishedAt":60,"linkOnly":61,"markdownContent":62,"markdownBody":63,"frontMatter":64,"refName":65,"commitSha":66,"skillDirectory":67,"forksCount":68,"repositoryHtmlUrl":69,"repositoryHomepageUrl":70,"ownerHtmlUrl":71,"ownerBlogUrl":72,"ownerDescription":73,"updatedAt":74,"install":75,"resources":88},"01M6WZZZHCTPZ8YF9564J3QJVT","waza-interactive",null,"Interactive workflow partner for creating, testing, and improving AI agent skills with waza. USE FOR: run my evals, check my skill, compare models, create eval suite, debug failing tests, is my skill ready, ship readiness, interpret results, improve score. DO NOT USE FOR: general coding, non-skill work, writing skill content (use skill-authoring), improving frontmatter only (use sensei).","microsoft\u002Fwaza","microsoft","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6154722?v=4","01M698VYDXZ6JW5WH3CRV650R8","Microsoft","01M6WZT4PAP160SPCF851532KH",1511,"MIT","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fskills\u002Fwaza-interactive\u002FSKILL.md",[18,23,28],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",{"categoryNo":29,"categoryName":30,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000036","内容创作者",[32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59],"waza","waza interactive","interactive","use","failing","general coding","eval suite","for","authoring","tests","do","general","evals","ready","only","testing","eval","work","improve","my skill","run","create","debug","sensei","create eval","workflow","ai agent","writing","2026-11-10 10:12:30",false,"---\nname: waza-interactive\ndescription: \"Interactive workflow partner for creating, testing, and improving AI agent skills with waza. USE FOR: run my evals, check my skill, compare models, create eval suite, debug failing tests, is my skill ready, ship readiness, interpret results, improve score. DO NOT USE FOR: general coding, non-skill work, writing skill content (use skill-authoring), improving frontmatter only (use sensei).\"\n---\n\n# Waza Interactive\n\nYou are a workflow partner that orchestrates waza evaluations conversationally. Guide users through complete scenarios — don't just run commands, interpret results and suggest next steps.\n\n## Available MCP Tools\n\nCall these tools to execute waza operations:\n\n| Tool | Purpose |\n|------|---------|\n| `waza_eval_list` | List available eval suites |\n| `waza_eval_get` | Get eval spec details |\n| `waza_eval_validate` | Validate eval YAML syntax |\n| `waza_eval_run` | Execute an eval benchmark |\n| `waza_task_list` | List tasks in an eval |\n| `waza_run_status` | Poll running eval status |\n| `waza_run_cancel` | Cancel a running eval |\n| `waza_results_summary` | Get aggregate scores |\n| `waza_results_runs` | Get per-task run details |\n| `waza_skill_check` | Check skill compliance |\n\n## Scenario 1: Create a New Eval\n\nWhen user wants to create an eval suite for their skill:\n\n1. Ask which skill to evaluate — get the skill name and path\n2. Call `waza_eval_list` to check for existing evals for this skill\n3. If none exist, run `waza init \u003Cdirectory>` via terminal to scaffold\n4. Explain the generated `eval.yaml` structure — name, skill, executor, tasks\n5. Help define tasks: ask what behaviors to test, suggest validators (`code`, `regex`)\n6. For each task, help write the prompt and expected output\n7. Call `waza_eval_validate` to confirm the YAML is valid\n8. Suggest running with `waza_eval_run` to verify the first task passes\n\n**Key guidance:** Start with 3–5 tasks covering happy path, edge case, and error handling.\n\n## Scenario 2: Run and Interpret Results\n\nWhen user wants to run evals and understand scores:\n\n1. Call `waza_eval_run` with the eval spec path and context dir\n2. Poll `waza_run_status` until complete (check every 10s)\n3. Call `waza_results_summary` to get aggregate scores\n4. Interpret the results for the user:\n   - **Pass rate** — percentage of tasks that passed all validators\n   - **Weighted score** — 0.0–1.0 aggregate across all tasks\n   - **Duration** — total and per-task execution time\n5. If pass rate \u003C 80%, identify which tasks failed and why\n6. Call `waza_results_runs` for per-task details on failures\n7. Suggest specific improvements: prompt rewording, validator tuning, fixture updates\n\n**Thresholds:** ≥90% pass rate = strong, 70–89% = needs work, \u003C70% = significant issues.\n\n## Scenario 3: Compare Models\n\nWhen user wants to compare model performance:\n\n1. Ask which models to compare (e.g., gpt-4o vs claude-sonnet-4)\n2. Call `waza_eval_run` with model A — save results\n3. Call `waza_eval_run` with model B — save results\n4. Compare results side by side:\n   - Per-task pass\u002Ffail differences\n   - Score deltas (which model scores higher on which tasks)\n   - Duration differences (speed vs quality tradeoff)\n5. Provide a recommendation: which model is better for this skill and why\n6. Suggest next steps: try a third model, tune prompts for the weaker model, or adjust validators\n\n**Guidance:** Run each model 2–3 times to account for variance before drawing conclusions.\n\n## Scenario 4: Debug a Failing Skill\n\nWhen user's skill is failing evals or behaving unexpectedly:\n\n1. Call `waza_skill_check` to verify skill compliance (frontmatter, triggers, token count)\n2. If compliance issues found, fix those first — they affect routing\n3. Call `waza_eval_run` with `--verbose` and `--transcript-dir` flags\n4. Call `waza_results_runs` to get per-task failure details\n5. Analyze failure patterns:\n   - **All tasks fail** → prompt or fixture issue, check skill instructions\n   - **Some tasks fail** → specific edge cases, review failed task prompts\n   - **Validator failures** → regex too strict, code validator language mismatch\n6. Suggest targeted fixes based on the pattern\n7. Re-run with `waza_eval_run` to verify the fix\n\n## Scenario 5: Ship Readiness Check\n\nWhen user asks \"is my skill ready?\" or wants a pre-ship checklist:\n\n1. Call `waza_skill_check` — verify compliance score ≥ medium-high\n2. Call `waza_eval_validate` — confirm eval YAML is valid\n3. Call `waza_eval_run` — execute full eval suite\n4. Call `waza_results_summary` — check aggregate scores\n5. Render the readiness verdict:\n\n```\nSHIP READINESS CHECKLIST:\n☐ Skill compliance: [score] (need: medium-high+)\n☐ Eval YAML valid: [yes\u002Fno]\n☐ Pass rate: [X]% (need: ≥90%)\n☐ Weighted score: [X.XX] (need: ≥0.85)\n☐ No task timeouts\n☐ Consistent across 2+ runs\n\nVERDICT: [READY \u002F NOT READY — fix items marked ✗]\n```\n\n6. If NOT READY, route to the appropriate scenario (Scenario 4 for failures, Scenario 1 for missing evals)\n\n## Conversation Style\n\n- Always explain *why* before *what* — context before commands\n- After every tool call, interpret the result in plain language\n- When something fails, diagnose before suggesting fixes\n- Offer the next logical step — don't wait to be asked\n- Use the checklist format for multi-step validations\n","# Waza Interactive\n\nYou are a workflow partner that orchestrates waza evaluations conversationally. Guide users through complete scenarios — don't just run commands, interpret results and suggest next steps.\n\n## Available MCP Tools\n\nCall these tools to execute waza operations:\n\n| Tool | Purpose |\n|------|---------|\n| `waza_eval_list` | List available eval suites |\n| `waza_eval_get` | Get eval spec details |\n| `waza_eval_validate` | Validate eval YAML syntax |\n| `waza_eval_run` | Execute an eval benchmark |\n| `waza_task_list` | List tasks in an eval |\n| `waza_run_status` | Poll running eval status |\n| `waza_run_cancel` | Cancel a running eval |\n| `waza_results_summary` | Get aggregate scores |\n| `waza_results_runs` | Get per-task run details |\n| `waza_skill_check` | Check skill compliance |\n\n## Scenario 1: Create a New Eval\n\nWhen user wants to create an eval suite for their skill:\n\n1. Ask which skill to evaluate — get the skill name and path\n2. Call `waza_eval_list` to check for existing evals for this skill\n3. If none exist, run `waza init \u003Cdirectory>` via terminal to scaffold\n4. Explain the generated `eval.yaml` structure — name, skill, executor, tasks\n5. Help define tasks: ask what behaviors to test, suggest validators (`code`, `regex`)\n6. For each task, help write the prompt and expected output\n7. Call `waza_eval_validate` to confirm the YAML is valid\n8. Suggest running with `waza_eval_run` to verify the first task passes\n\n**Key guidance:** Start with 3–5 tasks covering happy path, edge case, and error handling.\n\n## Scenario 2: Run and Interpret Results\n\nWhen user wants to run evals and understand scores:\n\n1. Call `waza_eval_run` with the eval spec path and context dir\n2. Poll `waza_run_status` until complete (check every 10s)\n3. Call `waza_results_summary` to get aggregate scores\n4. Interpret the results for the user:\n   - **Pass rate** — percentage of tasks that passed all validators\n   - **Weighted score** — 0.0–1.0 aggregate across all tasks\n   - **Duration** — total and per-task execution time\n5. If pass rate \u003C 80%, identify which tasks failed and why\n6. Call `waza_results_runs` for per-task details on failures\n7. Suggest specific improvements: prompt rewording, validator tuning, fixture updates\n\n**Thresholds:** ≥90% pass rate = strong, 70–89% = needs work, \u003C70% = significant issues.\n\n## Scenario 3: Compare Models\n\nWhen user wants to compare model performance:\n\n1. Ask which models to compare (e.g., gpt-4o vs claude-sonnet-4)\n2. Call `waza_eval_run` with model A — save results\n3. Call `waza_eval_run` with model B — save results\n4. Compare results side by side:\n   - Per-task pass\u002Ffail differences\n   - Score deltas (which model scores higher on which tasks)\n   - Duration differences (speed vs quality tradeoff)\n5. Provide a recommendation: which model is better for this skill and why\n6. Suggest next steps: try a third model, tune prompts for the weaker model, or adjust validators\n\n**Guidance:** Run each model 2–3 times to account for variance before drawing conclusions.\n\n## Scenario 4: Debug a Failing Skill\n\nWhen user's skill is failing evals or behaving unexpectedly:\n\n1. Call `waza_skill_check` to verify skill compliance (frontmatter, triggers, token count)\n2. If compliance issues found, fix those first — they affect routing\n3. Call `waza_eval_run` with `--verbose` and `--transcript-dir` flags\n4. Call `waza_results_runs` to get per-task failure details\n5. Analyze failure patterns:\n   - **All tasks fail** → prompt or fixture issue, check skill instructions\n   - **Some tasks fail** → specific edge cases, review failed task prompts\n   - **Validator failures** → regex too strict, code validator language mismatch\n6. Suggest targeted fixes based on the pattern\n7. Re-run with `waza_eval_run` to verify the fix\n\n## Scenario 5: Ship Readiness Check\n\nWhen user asks \"is my skill ready?\" or wants a pre-ship checklist:\n\n1. Call `waza_skill_check` — verify compliance score ≥ medium-high\n2. Call `waza_eval_validate` — confirm eval YAML is valid\n3. Call `waza_eval_run` — execute full eval suite\n4. Call `waza_results_summary` — check aggregate scores\n5. Render the readiness verdict:\n\n```\nSHIP READINESS CHECKLIST:\n☐ Skill compliance: [score] (need: medium-high+)\n☐ Eval YAML valid: [yes\u002Fno]\n☐ Pass rate: [X]% (need: ≥90%)\n☐ Weighted score: [X.XX] (need: ≥0.85)\n☐ No task timeouts\n☐ Consistent across 2+ runs\n\nVERDICT: [READY \u002F NOT READY — fix items marked ✗]\n```\n\n6. If NOT READY, route to the appropriate scenario (Scenario 4 for failures, Scenario 1 for missing evals)\n\n## Conversation Style\n\n- Always explain *why* before *what* — context before commands\n- After every tool call, interpret the result in plain language\n- When something fails, diagnose before suggesting fixes\n- Offer the next logical step — don't wait to be asked\n- Use the checklist format for multi-step validations\n",{"name":5,"description":7},"main","116214a33b5e07766539f64c0ea21521467829bf","skills\u002Fwaza-interactive",101,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza","https:\u002F\u002Fmicrosoft.github.io\u002Fwaza\u002F","https:\u002F\u002Fgithub.com\u002Fmicrosoft","https:\u002F\u002Fopensource.microsoft.com","Open source projects and samples from Microsoft","2026-11-15 04:18:35",{"source":8,"skillName":5,"commands":76,"prompt":86,"zipAvailable":87},[77,80,83],{"tool":78,"command":79},"npx","npx skills add microsoft\u002Fwaza --skill waza-interactive",{"tool":81,"command":82},"bunx","bunx skills add microsoft\u002Fwaza --skill waza-interactive",{"tool":84,"command":85},"pnpm","pnpm dlx skills add microsoft\u002Fwaza --skill waza-interactive","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fskills\u002Fwaza-interactive\u002FSKILL.md",true,[89],{"path":90,"resourceType":91,"resourceTypeDesc":92,"sizeBytes":93,"sourceHtmlUrl":94},"tests\u002Feval.yaml","REFERENCE","参考文档",2300,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fskills\u002Fwaza-interactive\u002Ftests\u002Feval.yaml",[96,117,139,159,172,183],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":106,"publishedAt":60},"01M6X00A438VG96V20V1VMDHCM","waza-runner","Run evaluations on Agent Skills to measure their effectiveness. \nUSE FOR: \"run skill evals\", \"evaluate my skill\", \"test skill quality\", \n\"check skill triggers\", \"skill compliance check\", \"measure skill performance\",\n\"run evals on [skill-name]\", \"grade skill execution\".\nDO NOT USE FOR: writing skills (use skill-authoring), improving frontmatter \n(use sensei), or general testing unrelated to skills.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fwaza-runner\u002FSKILL.md",[102,105],{"categoryNo":103,"categoryName":104,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000008","测试与质量",{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[107,108,32,39,109,110,42,43,44,111,112,51,52,55,113,114,59,47,40,115,116],"waza runner","runner","frontmatter","run evals","unrelated","writing skills","Test Skill","performance","triggers","my",{"skillNo":118,"name":32,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":120,"categories":121,"keywords":124,"publishedAt":60},"01M6X003ZZKY5WQ5SWPAQ3MJMS","**WORKFLOW SKILL** - Evaluate AI agent skills using structured benchmarks with YAML specs, fixture isolation, and pluggable validators. USE FOR: run waza, waza help, run eval, run benchmark, evaluate skill, test agent, generate eval suite, init eval, compare results, score agent, agent evaluation, skill testing, cross-model comparison. DO NOT USE FOR: improving skill frontmatter (use waza dev), creating new skills from scratch (use skill-creator), token counting or budget checks (use waza tokens). INVOKES: Copilot SDK executor, mock engine, code\u002Fregex validators. FOR SINGLE OPERATIONS: use waza run directly for a single benchmark.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fskills\u002Fwaza\u002FSKILL.md",[122,123],{"categoryNo":103,"categoryName":104,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[32,125,126,127,128,48,129,130,131,132,133,52,134,135,136,57,137,138,38,47,58],"compare results","workflow skill","executor","fixture","skill-creator","model","Evaluation","evaluate skill","code","generate","copilot","init","counting","new",{"skillNo":140,"name":141,"nameZh":6,"description":142,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":143,"categories":144,"keywords":146,"publishedAt":60},"01M6WZZTQR955SVJ02J33DGQYK","code-explainer","Explains code snippets in plain English, breaking down what the code does step by step. Perfect for learning, code reviews, or documentation.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fskills\u002Fcode-explainer\u002FSKILL.md",[145],{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[147,148,133,149,150,151,152,39,153,154,155,156,157,158],"code explainer","explainer","step by step","plain english","plain","down","reviews","english","snippets","what the","learning","documentation",{"skillNo":160,"name":161,"nameZh":6,"description":162,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":163,"categories":164,"keywords":165,"publishedAt":60},"01M6WZZR27J4RE4J10M38ECM26","compatibility-skill","Read fixture status and report whether it is ready.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Finternal\u002Ftestdata\u002Fcompatibility\u002Fv1\u002FSKILL.md",[],[166,167,168,128,45,169,170,171],"compatibility skill","compatibility","it","read","report","status",{"skillNo":173,"name":174,"nameZh":6,"description":175,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":176,"categories":177,"keywords":178,"publishedAt":60},"01M6WZZMYWAKWE36WD4PHQK8V2","evidence-judge","Read a supplied report and summarize its recorded status and numeric items.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fexamples\u002Fevidence-judge\u002FSKILL.md",[],[179,180,181,171,182,170,169],"evidence judge","judge","evidence","summarize",{"skillNo":184,"name":141,"nameZh":6,"description":185,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":186,"categories":187,"keywords":192,"publishedAt":60},"01M6WZZE4CBXD08ZXA3SGPK2BF","**UTILITY SKILL** - Explain code snippets, functions, and algorithms in plain language.\nUSE FOR: explain code, what does this code do, break down this function,\nhelp me understand this, walk through this algorithm, clarify this logic,\nexplain this snippet, describe what happens here.\nDO NOT USE FOR: writing new code (use code generation), fixing bugs (use debugging),\nrefactoring (use refactoring skills), code review with action items.\nINVOKES: file reading tools to access code, language detection for tailored explanations.\nFOR SINGLE OPERATIONS: If the user just needs to see file contents, use file reading tools directly.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fwaza\u002Fblob\u002Fmain\u002Fexamples\u002Fcode-explainer\u002FSKILL.md",[188,191],{"categoryNo":189,"categoryName":190,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000007","代码审查",{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},[147,133,148,155,138,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,59,208],"action items","describe","file reading","break","action","reading","utility","me","clarify","access","refactoring","Code Generation","explain code","debugging","review","code review"]