[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EZCDFVR7F5K77S5WZ8RYJ5":3,"skill-same:01M7EZCDFVR7F5K77S5WZ8RYJ5":52},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7EZCDFVR7F5K77S5WZ8RYJ5","domain-agent-coaching",null,"Coach the design and ongoing reflection of long-lived domain Agents and Agent organizations. Use when defining or revising Identity, Domain, Scope, Organization or Collaboration relationships; deciding whether recurring work should become an Agent, Internal Agent, Skill, or Task; examining overlap, gaps, overloaded domains, unclear external roles, or unhealthy collaboration paths. Guide with evidence and questions without assuming the business context or making the owner's decisions.","lza6\u002FArmor-piercing-system-prompts","lza6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F128880206?v=4","01M7EB3K7WPX5T28EJ2NVHE6EG","Chinese-tingfeng","01M7EB3K97E1KQ16QWKJMFNGX8",11,"MIT","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fdomain-agent-coaching\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: domain-agent-coaching\ndescription: \"Coach the design and ongoing reflection of long-lived domain Agents and Agent organizations. Use when defining or revising Identity, Domain, Scope, Organization or Collaboration relationships; deciding whether recurring work should become an Agent, Internal Agent, Skill, or Task; examining overlap, gaps, overloaded domains, unclear external roles, or unhealthy collaboration paths. Guide with evidence and questions without assuming the business context or making the owner's decisions.\"\n---\n\n# Domain Agent Coaching\n\nHelp people and Agents reason about durable Agent design. Act as a coach: reveal assumptions, sharpen boundaries, gather perspectives, and leave consequential business decisions with their owner.\n\n## Hold the durable-Agent frame\n\nA Domain Agent is a long-lived working subject, not a task wrapper. Its Profile defines a stable Identity, Domain, and Scope; its Codex Thread retains trajectory and working context across many interactions.\n\nKeep these concepts separate:\n\n- **Agent**: a durable domain with its own continuity, responsibilities, and relationships.\n- **Internal Agent**: a durable subdomain that primarily serves one parent Agent and is normally reached through that parent.\n- **Skill**: reusable procedural knowledge or a capability an Agent can invoke. It does not own a domain.\n- **Task**: a bounded outcome performed by an Agent. It does not justify a durable identity by itself.\n\nDo not recommend a new Agent merely because work is large. Look for recurring responsibility, independent context, stable boundaries, and a continuing need to coordinate.\n\n## Build an evidence base\n\nInspect the current state before coaching:\n\n```sh\nloom team\nloom team AGENT\nloom agent get AGENT\nloom profile get AGENT\n```\n\nUse Profiles as declared intent, Organization and Collaboration as declared structure, and message Activity as behavioral evidence. None is complete alone.\n\nLabel what you learn:\n\n- **Fact**: explicitly declared or directly observed.\n- **Signal**: repeated behavior that may indicate a design issue.\n- **Inference**: a plausible interpretation that needs validation.\n- **Decision**: a choice confirmed by the responsible human or Agent owner.\n\nDo not turn message volume into authority, or a temporary project spike into a permanent scope boundary.\n\n## Use workload evidence on demand\n\nRead workload signals only after a user, Owner, incident, repeated delay, reported capacity problem, or planned\norganization review has created a concrete coaching question. Do not scan all Agents routinely or treat the output\nas a ranking.\n\nStart with the organization and then inspect the affected Agent, unless a specific Message, Inbox item, or Turn is\nalready known:\n\n```sh\nloom workload --days 7\nloom workload AGENT --days 1 --evidence\nloom workload AGENT --days 30\nloom workload AGENT --days 30 --evidence\nloom workload AGENT --days 90 --json\n```\n\nInterpret the report as a question-forming evidence source:\n\n- `calendar non-executing proxy` is not online idle time; it includes machine and service downtime.\n- Queue percentiles are meaningful only with their sample count. `n=0` means no completed sample, not zero wait.\n- Read executing, queue wait, current backlog, work source, wait reason, and data quality together.\n- Use stable evidence IDs to inspect only the context needed for the active coaching question.\n- Do not infer overloaded, underused, split, or merge from a percentage alone.\n\nHigh wait may come from capacity, an active Goal reservation, restart drain, Connector behavior, or scheduling. Low\nexecution may be healthy for a sparse high-risk Domain. Form alternative explanations, then interview the relevant\nOwner and Agent before proposing an organization change.\n\n## Run the coaching loop\n\n1. **Frame the tension.** State the observed design question without prescribing an answer.\n2. **Inspect the structure.** Read the affected Profiles, relationships, and relevant communication evidence.\n3. **Interview perspectives.** Ask the owner and affected Agents what they believe they own, depend on, reject, and repeatedly reconstruct.\n4. **Test the boundaries.** Look for overlap, orphaned decisions, hidden routing, overloaded context, duplicated judgment, or an external role conflicting with internal responsibility.\n5. **Offer alternatives.** Present a small number of viable designs and their tradeoffs.\n6. **Name the decision owner.** Make clear who understands the business well enough to decide.\n7. **Record only confirmed changes.** Do not mutate Profiles or relationships without explicit confirmation.\n8. **Revisit after evidence accumulates.** Organization design is continuous; one conversation is not proof that a structure works.\n\nUse `$loom-communication` when gathering another Agent's perspective. Ask about that Agent's own experience; do not ask it to decide the whole organization.\n\n## Questions that expose boundaries\n\nUse only the questions relevant to the tension:\n\n- What durable judgment should this Agent become better at over months?\n- What enters this Domain, and what must leave it?\n- Which decisions can this Agent make without rebuilding context or consulting another owner?\n- Which requests repeatedly arrive but do not belong here?\n- What knowledge is reusable procedure, and what responsibility requires a continuing subject?\n- If this Domain is split, what context and decisions remain with the parent?\n- Who may contact an Internal Agent directly, and why?\n- Does an external channel role express this Agent's Domain, or require a dedicated Interface Agent?\n- What evidence would show that the proposed structure is wrong?\n\n## Produce a decision artifact\n\nKeep the output concise and reviewable:\n\n```markdown\n## Signal\nObserved facts and repeated behavior.\n\n## Design Question\nThe boundary or organizational choice that needs examination.\n\n## Perspectives\nWhat affected people and Agents believe, with disagreements preserved.\n\n## Unresolved Assumptions\nWhat is still inferred rather than known.\n\n## Possible Designs\nTwo or three options with consequences and reversibility.\n\n## Decision Owner\nWho has enough business context and authority to choose.\n\n## Decision\nConfirmed choice, or `open` when no choice has been made.\n```\n\nNever manufacture consensus. If evidence is insufficient, leave the decision open and identify the next observation or conversation that would reduce uncertainty.\n","# Domain Agent Coaching\n\nHelp people and Agents reason about durable Agent design. Act as a coach: reveal assumptions, sharpen boundaries, gather perspectives, and leave consequential business decisions with their owner.\n\n## Hold the durable-Agent frame\n\nA Domain Agent is a long-lived working subject, not a task wrapper. Its Profile defines a stable Identity, Domain, and Scope; its Codex Thread retains trajectory and working context across many interactions.\n\nKeep these concepts separate:\n\n- **Agent**: a durable domain with its own continuity, responsibilities, and relationships.\n- **Internal Agent**: a durable subdomain that primarily serves one parent Agent and is normally reached through that parent.\n- **Skill**: reusable procedural knowledge or a capability an Agent can invoke. It does not own a domain.\n- **Task**: a bounded outcome performed by an Agent. It does not justify a durable identity by itself.\n\nDo not recommend a new Agent merely because work is large. Look for recurring responsibility, independent context, stable boundaries, and a continuing need to coordinate.\n\n## Build an evidence base\n\nInspect the current state before coaching:\n\n```sh\nloom team\nloom team AGENT\nloom agent get AGENT\nloom profile get AGENT\n```\n\nUse Profiles as declared intent, Organization and Collaboration as declared structure, and message Activity as behavioral evidence. None is complete alone.\n\nLabel what you learn:\n\n- **Fact**: explicitly declared or directly observed.\n- **Signal**: repeated behavior that may indicate a design issue.\n- **Inference**: a plausible interpretation that needs validation.\n- **Decision**: a choice confirmed by the responsible human or Agent owner.\n\nDo not turn message volume into authority, or a temporary project spike into a permanent scope boundary.\n\n## Use workload evidence on demand\n\nRead workload signals only after a user, Owner, incident, repeated delay, reported capacity problem, or planned\norganization review has created a concrete coaching question. Do not scan all Agents routinely or treat the output\nas a ranking.\n\nStart with the organization and then inspect the affected Agent, unless a specific Message, Inbox item, or Turn is\nalready known:\n\n```sh\nloom workload --days 7\nloom workload AGENT --days 1 --evidence\nloom workload AGENT --days 30\nloom workload AGENT --days 30 --evidence\nloom workload AGENT --days 90 --json\n```\n\nInterpret the report as a question-forming evidence source:\n\n- `calendar non-executing proxy` is not online idle time; it includes machine and service downtime.\n- Queue percentiles are meaningful only with their sample count. `n=0` means no completed sample, not zero wait.\n- Read executing, queue wait, current backlog, work source, wait reason, and data quality together.\n- Use stable evidence IDs to inspect only the context needed for the active coaching question.\n- Do not infer overloaded, underused, split, or merge from a percentage alone.\n\nHigh wait may come from capacity, an active Goal reservation, restart drain, Connector behavior, or scheduling. Low\nexecution may be healthy for a sparse high-risk Domain. Form alternative explanations, then interview the relevant\nOwner and Agent before proposing an organization change.\n\n## Run the coaching loop\n\n1. **Frame the tension.** State the observed design question without prescribing an answer.\n2. **Inspect the structure.** Read the affected Profiles, relationships, and relevant communication evidence.\n3. **Interview perspectives.** Ask the owner and affected Agents what they believe they own, depend on, reject, and repeatedly reconstruct.\n4. **Test the boundaries.** Look for overlap, orphaned decisions, hidden routing, overloaded context, duplicated judgment, or an external role conflicting with internal responsibility.\n5. **Offer alternatives.** Present a small number of viable designs and their tradeoffs.\n6. **Name the decision owner.** Make clear who understands the business well enough to decide.\n7. **Record only confirmed changes.** Do not mutate Profiles or relationships without explicit confirmation.\n8. **Revisit after evidence accumulates.** Organization design is continuous; one conversation is not proof that a structure works.\n\nUse `$loom-communication` when gathering another Agent's perspective. Ask about that Agent's own experience; do not ask it to decide the whole organization.\n\n## Questions that expose boundaries\n\nUse only the questions relevant to the tension:\n\n- What durable judgment should this Agent become better at over months?\n- What enters this Domain, and what must leave it?\n- Which decisions can this Agent make without rebuilding context or consulting another owner?\n- Which requests repeatedly arrive but do not belong here?\n- What knowledge is reusable procedure, and what responsibility requires a continuing subject?\n- If this Domain is split, what context and decisions remain with the parent?\n- Who may contact an Internal Agent directly, and why?\n- Does an external channel role express this Agent's Domain, or require a dedicated Interface Agent?\n- What evidence would show that the proposed structure is wrong?\n\n## Produce a decision artifact\n\nKeep the output concise and reviewable:\n\n```markdown\n## Signal\nObserved facts and repeated behavior.\n\n## Design Question\nThe boundary or organizational choice that needs examination.\n\n## Perspectives\nWhat affected people and Agents believe, with disagreements preserved.\n\n## Unresolved Assumptions\nWhat is still inferred rather than known.\n\n## Possible Designs\nTwo or three options with consequences and reversibility.\n\n## Decision Owner\nWho has enough business context and authority to choose.\n\n## Decision\nConfirmed choice, or `open` when no choice has been made.\n```\n\nNever manufacture consensus. If evidence is insufficient, leave the decision open and identify the next observation or conversation that would reduce uncertainty.\n",{"name":5,"description":7},"main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","05-技能与代理开发\u002Fdomain-agent-coaching",3,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts","https:\u002F\u002Fgithub.com\u002Flza6","个人vx：Tf00798\r\n个人✈：@TFGY999","2026-10-08 14:29:22",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill domain-agent-coaching",{"tool":38,"command":39},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill domain-agent-coaching",{"tool":41,"command":42},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill domain-agent-coaching","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fdomain-agent-coaching\u002FSKILL.md",true,[46],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"agents\u002Fopenai.yaml","REFERENCE","参考文档",246,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fdomain-agent-coaching\u002Fagents\u002Fopenai.yaml",[53,60,65,72,79,86],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M7FFG485KY83TD6HSC4YPCMC","obsidian-clipper-template-creator","Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator\u002FSKILL.md",[],[],{"skillNo":61,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M7FFG0T6AZ0STEJPSB3CX84K","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator-antigravity-awesome-skills-main\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7FFFY0QEM14SKWPDYGMXM2P","python-api","Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP\u002FmAP semantics, and source-style import diagnostics.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Fpython-api\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M7FFFVZQZ2GJ0C0TGP7JHD20","file-evaluation","Folder-based PASCAL VOC AP\u002FmAP evaluation from ground-truth and detection text files.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Ffile-evaluation\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7FFFSFNYSE77SJ6PE1CK2DA","object-detection-metrics","Guides legacy Object-Detection-Metrics PASCAL VOC AP\u002FmAP evaluation from detection text folders and direct Python API classes.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M7FFFP0GCK1RVXBCPQ8APBMT","nx-workspace-patterns","Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fnx-workspace-patterns\u002FSKILL.md",[],[]]