[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C4792KBN8V7FEWB0RY04BV":3,"skill-same:01M7C4792KBN8V7FEWB0RY04BV":63},{"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":43,"commitSha":44,"skillDirectory":45,"forksCount":13,"repositoryHtmlUrl":46,"repositoryHomepageUrl":6,"ownerHtmlUrl":47,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":48,"install":49,"resources":62},"01M7C4792KBN8V7FEWB0RY04BV","grill-me",null,"Interview a plan when the user asks to be grilled.","fengwhang\u002Fmercury","fengwhang","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F325040254?v=4","01M7C3K54Y41KVGPHEMKEP9X70","01M7C3K55NTZ4B36YP2JY1NMJ0",0,"MIT","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fhermes\u002Fskills\u002Fproductivity\u002Fgrill-me\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: grill-me\ndescription: \"Interview a plan when the user asks to be grilled.\"\nversion: 3.0.0\nauthor: \"Rafael Zendron (rafaumeu) + Matt Pocock (mattpocock\u002Fskills, grilling) + Hermes Agent\"\nlicense: MIT\nplatforms: [linux, macos, windows]\nmetadata:\n  mercury:\n    tags: [planning, adversarial, interview, decision-tree, pre-implementation, review, alignment]\n    related_skills: [requesting-code-review, subagent-driven-development, test-driven-development]\n---\n\n# Grill Me\n\n## Mercury engine tools\n\nThis workflow is shared by Mercury Hermes and Mercury OMP. Tool-call examples\nusing `terminal`, `read_file`, or `delegate_task` are Hermes-shaped examples;\non OMP use the available `bash`, `read`, search, and `task` tools with their\nactual schemas. Never invoke an unavailable tool. For the complete mapping,\nload the `mercury-agent` skill and its `references\u002Fengine-tools.md`. Apply this\nmapping to the examples below; keep each engine's permissions unchanged.\n\nStress-tests a plan through structured adversarial questioning before any\ncode is written. Models the plan as a **design tree** — every decision\nbranches into the decisions that hang off it — and interviews the user in\nrounds until every branch is resolved and nothing is silently assumed.\n\nCombines the phase discipline of the original with the frontier-rounds\nmechanic from mattpocock\u002Fskills' `grilling`.\n\n## When to Use\n\n- User says \"grill me\", \"interview my plan\", \"stress test this idea\"\n\nUse this interview when requested. Do not interrupt an already authorized\nimplementation merely because its plan could be explored further.\n\n## Prerequisites\n\nNone. The skill works on any plan or raw idea.\n\n## Core Mechanic: Frontier Rounds\n\nMap the plan as a design tree. The **frontier** is every decision whose\nprerequisites are already settled — the questions you can ask NOW without\nguessing at answers you haven't heard yet.\n\nWork in **rounds**: ask the whole current frontier in one message, numbered,\neach question carrying your recommended answer. Then wait. A question whose\nanswer depends on another question still open in this round belongs to a\nLATER round, not this one.\n\nFormat each round like so:\n\n```\n❓ Q1 — \u003Cquestion title>: \u003Cquestion body, options if relevant>\n➡️ Recommendation: \u003Cyour recommended answer + one-line why>\n\n❓ Q2 — \u003Cquestion title>: \u003Cquestion body>\n➡️ Recommendation: \u003C...>\n```\n\nEach answer reshapes the tree: settled decisions push the frontier outward\nand unblock dependent questions. Recompute the frontier and ask the next\nround.\n\n**Facts are your job; decisions are the user's.** When a frontier question\nneeds a fact from the environment (codebase, filesystem, config, docs), find\nit yourself using the active engine's file, search, and shell tools — or\ndispatch a subagent through its advertised delegation tool for a heavy exploration. Never ask the user for\nanything you could look up. Don't block on an exploration: only the questions\ndownstream of it wait; ask the rest of the frontier now.\n\n## Question Coverage (work these branches into the tree)\n\n**Understanding** — the real goal and boundaries:\n- What is the ACTUAL objective? What is explicitly IN and OUT of scope?\n- What are the constraints (time, tech, team, budget)? Who are the users?\n\n**Technical decisions** — for each architectural choice:\n- \"Why this approach and not X?\" \u002F \"What happens if Y fails?\"\n- \"What's the worst case?\" \u002F \"How would you roll back?\"\n- Cross-reference the existing codebase; if the project already has a\n  pattern for this, call it out.\n\n**Edge cases:**\n- \"What happens if the user does Z?\" \u002F \"What if dependency X goes down?\"\n- \"What if volume is 100x expected?\" \u002F \"What are the security implications?\"\n\n## Synthesis (when the frontier is empty)\n\n1. Summarize the settled decisions, open items, and scope.\n2. Confirm alignment when the user has asked only for an interview.\n3. If implementation was already authorized, proceed within that scope once\n   required decisions are settled; do not ask for the same authorization again.\n\n## Pitfalls\n\n1. **Asking questions out of dependency order.** A question that depends on\n   an unanswered question is a guess wearing a question mark. Keep it for a\n   later round.\n2. **Skipping the codebase.** Find facts in code with Mercury tools instead of\n   asking the user.\n3. **Accepting \"I don't know\" as final.** Suggest options, explain\n   trade-offs, make a recommendation.\n4. **Losing the requested mode.** An interview produces a conversation, not\n   unsolicited files. Preserve any existing authorization to implement afterward.\n5. **Being too agreeable.** Your job is to find problems. If everything looks\n   fine, look harder.\n6. **Not adapting to the user's language.** Interview in whatever language\n   the user speaks.\n\n## Verification\n\n- [ ] Every question in a round had all its prerequisites already settled\n- [ ] Provided a recommendation with each question\n- [ ] Explored the codebase for facts instead of asking the user\n- [ ] Frontier empty (no branch silently assumed) before synthesizing\n- [ ] Produced a clear summary of all decisions and open items\n- [ ] Confirmed user alignment before stopping\n\nConsolidates Mercury's former `\u002Fgrilling` skill and `\u002Fgrill-me` wrapper.\nUse the current engine's available tool names for inspection and delegation.\n","# Grill Me\n\n## Mercury engine tools\n\nThis workflow is shared by Mercury Hermes and Mercury OMP. Tool-call examples\nusing `terminal`, `read_file`, or `delegate_task` are Hermes-shaped examples;\non OMP use the available `bash`, `read`, search, and `task` tools with their\nactual schemas. Never invoke an unavailable tool. For the complete mapping,\nload the `mercury-agent` skill and its `references\u002Fengine-tools.md`. Apply this\nmapping to the examples below; keep each engine's permissions unchanged.\n\nStress-tests a plan through structured adversarial questioning before any\ncode is written. Models the plan as a **design tree** — every decision\nbranches into the decisions that hang off it — and interviews the user in\nrounds until every branch is resolved and nothing is silently assumed.\n\nCombines the phase discipline of the original with the frontier-rounds\nmechanic from mattpocock\u002Fskills' `grilling`.\n\n## When to Use\n\n- User says \"grill me\", \"interview my plan\", \"stress test this idea\"\n\nUse this interview when requested. Do not interrupt an already authorized\nimplementation merely because its plan could be explored further.\n\n## Prerequisites\n\nNone. The skill works on any plan or raw idea.\n\n## Core Mechanic: Frontier Rounds\n\nMap the plan as a design tree. The **frontier** is every decision whose\nprerequisites are already settled — the questions you can ask NOW without\nguessing at answers you haven't heard yet.\n\nWork in **rounds**: ask the whole current frontier in one message, numbered,\neach question carrying your recommended answer. Then wait. A question whose\nanswer depends on another question still open in this round belongs to a\nLATER round, not this one.\n\nFormat each round like so:\n\n```\n❓ Q1 — \u003Cquestion title>: \u003Cquestion body, options if relevant>\n➡️ Recommendation: \u003Cyour recommended answer + one-line why>\n\n❓ Q2 — \u003Cquestion title>: \u003Cquestion body>\n➡️ Recommendation: \u003C...>\n```\n\nEach answer reshapes the tree: settled decisions push the frontier outward\nand unblock dependent questions. Recompute the frontier and ask the next\nround.\n\n**Facts are your job; decisions are the user's.** When a frontier question\nneeds a fact from the environment (codebase, filesystem, config, docs), find\nit yourself using the active engine's file, search, and shell tools — or\ndispatch a subagent through its advertised delegation tool for a heavy exploration. Never ask the user for\nanything you could look up. Don't block on an exploration: only the questions\ndownstream of it wait; ask the rest of the frontier now.\n\n## Question Coverage (work these branches into the tree)\n\n**Understanding** — the real goal and boundaries:\n- What is the ACTUAL objective? What is explicitly IN and OUT of scope?\n- What are the constraints (time, tech, team, budget)? Who are the users?\n\n**Technical decisions** — for each architectural choice:\n- \"Why this approach and not X?\" \u002F \"What happens if Y fails?\"\n- \"What's the worst case?\" \u002F \"How would you roll back?\"\n- Cross-reference the existing codebase; if the project already has a\n  pattern for this, call it out.\n\n**Edge cases:**\n- \"What happens if the user does Z?\" \u002F \"What if dependency X goes down?\"\n- \"What if volume is 100x expected?\" \u002F \"What are the security implications?\"\n\n## Synthesis (when the frontier is empty)\n\n1. Summarize the settled decisions, open items, and scope.\n2. Confirm alignment when the user has asked only for an interview.\n3. If implementation was already authorized, proceed within that scope once\n   required decisions are settled; do not ask for the same authorization again.\n\n## Pitfalls\n\n1. **Asking questions out of dependency order.** A question that depends on\n   an unanswered question is a guess wearing a question mark. Keep it for a\n   later round.\n2. **Skipping the codebase.** Find facts in code with Mercury tools instead of\n   asking the user.\n3. **Accepting \"I don't know\" as final.** Suggest options, explain\n   trade-offs, make a recommendation.\n4. **Losing the requested mode.** An interview produces a conversation, not\n   unsolicited files. Preserve any existing authorization to implement afterward.\n5. **Being too agreeable.** Your job is to find problems. If everything looks\n   fine, look harder.\n6. **Not adapting to the user's language.** Interview in whatever language\n   the user speaks.\n\n## Verification\n\n- [ ] Every question in a round had all its prerequisites already settled\n- [ ] Provided a recommendation with each question\n- [ ] Explored the codebase for facts instead of asking the user\n- [ ] Frontier empty (no branch silently assumed) before synthesizing\n- [ ] Produced a clear summary of all decisions and open items\n- [ ] Confirmed user alignment before stopping\n\nConsolidates Mercury's former `\u002Fgrilling` skill and `\u002Fgrill-me` wrapper.\nUse the current engine's available tool names for inspection and delegation.\n",{"name":5,"author":23,"license":14,"version":24,"metadata":25,"platforms":39,"description":7},"Rafael Zendron (rafaumeu) + Matt Pocock (mattpocock\u002Fskills, grilling) + Hermes Agent","3.0.0",{"mercury":26},{"tags":27,"related_skills":35},[28,29,30,31,32,33,34],"planning","adversarial","interview","decision-tree","pre-implementation","review","alignment",[36,37,38],"requesting-code-review","subagent-driven-development","test-driven-development",[40,41,42],"linux","macos","windows","main","94a12cab2ce32e2020d179b5d579d0cbcbd0f206","hermes\u002Fskills\u002Fproductivity\u002Fgrill-me","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury","https:\u002F\u002Fgithub.com\u002Ffengwhang","2026-10-09 08:26:42",{"source":8,"skillName":5,"commands":50,"prompt":60,"zipAvailable":61},[51,54,57],{"tool":52,"command":53},"npx","npx skills add fengwhang\u002Fmercury --skill grill-me",{"tool":55,"command":56},"bunx","bunx skills add fengwhang\u002Fmercury --skill grill-me",{"tool":58,"command":59},"pnpm","pnpm dlx skills add fengwhang\u002Fmercury --skill grill-me","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fhermes\u002Fskills\u002Fproductivity\u002Fgrill-me\u002FSKILL.md",true,[],[64,71,78,85,91,98],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":18},"01M7C4BTY3ND38CN200ZCTH6NV","greet","A greeting skill for testing","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Fmarketplace\u002Ffixtures\u002Fvalid-marketplace\u002Fplugins\u002Fhello-plugin\u002Fskills\u002Fgreet\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":18},"01M7C4BSNZTPW8JR5NGCQXGSEZ","valid-skill","A valid skill for testing purposes.","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Ffixtures\u002Fskills\u002Fvalid-skill\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":18},"01M7C4BRA7J8C06KSVW100F1W1","unknown-field","A skill with an unknown frontmatter field.","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Ffixtures\u002Fskills\u002Funknown-field\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":18},"01M7C4BNB04KQGCJ835JE8AR13","no-frontmatter","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Ffixtures\u002Fskills\u002Fno-frontmatter\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":18},"01M7C4BM38TZFYT1AJ6SKHPZZ2","child-skill","A nested skill in a subdirectory.","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Ffixtures\u002Fskills\u002Fnested\u002Fchild-skill\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":18},"01M7C4BJDN7RFFCM28CGX15Q97","different-name","A skill with a name that doesn't match the directory.","https:\u002F\u002Fgithub.com\u002Ffengwhang\u002Fmercury\u002Fblob\u002Fmain\u002Fomp\u002Fpackages\u002Fcoding-agent\u002Ftest\u002Ffixtures\u002Fskills\u002Fname-mismatch\u002FSKILL.md",[],[]]