[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F6C1RY3KEX7MZ34NF0SR2V":3,"skill-same:01M7F6C1RY3KEX7MZ34NF0SR2V":71},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"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":23,"commitSha":24,"skillDirectory":25,"forksCount":13,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":28,"ownerDescription":29,"updatedAt":30,"install":31,"resources":44},"01M7F6C1RY3KEX7MZ34NF0SR2V","derailment-method",null,"yigitkonur\u002Fplugin-agentic-maker","yigitkonur","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9989650?v=4","01M7548A59WS0A38NSK8FASK46","Yigit Konur","01M7F6AQQPM8KBP08NQ3BG7160",0,"MIT","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: derailment-method\ndescription: How to find why an agent drifted while using a skill, CLI or tool: read the whole run, name the text that misled it, rewrite that text. Use when diagnosing \"why did my agent skip \u002F guess \u002F ignore this\". Not for code review.\nuser-invocable: false\n---\n\n\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#Audit Skill by Derailment: The Native Two-Tier Auditor -->\n# Derailment method\n\nA skill, a tool description, a `--help` page and an error message are a program that an LLM executes. When the agent drifts, the defect is in the text or affordance it was given. This method finds that text from the evidence of a real run and rewrites it. Two tiers do the work: a **worker** (a fresh agent that attempts a realistic task) and a **reader** (a read-only agent that studies the worker's whole execution). The orchestrator applies fixes and reruns.\n\n\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#The 5 Non-Negotiable Invariants -->\n## The five invariants\n\n1. **Fix the text, not the executor.** Every remedy is an edit to the controlling text (a skill sentence, an example, a routing row, a tool description, an error message, a script) or to the tool itself. Never conclude \"use a smarter agent\". Rewrite or delete the sentence that misled; do not leave it and add a \"do \u002F don't\" rule beside it.\n2. **Claims are leads; the machine is the witness.** An agent's closing report is a claim. Verify exit codes and what is on disk (`agentic-maker check receipts`) before believing \"done\". See [evidence-and-anti-sycophancy](references\u002Fevidence-and-anti-sycophancy.md).\n3. **Never judge from a view that hides what the agent did.** Read the harness's structured record (every tool call with full arguments, every result, hook events). A folded terminal view, a summary or the agent's own account is not evidence. Source ranking: `skills\u002Fderail\u002Freferences\u002Ftrace-sources.md`.\n4. **Search by dimension; do not chase a quota.** Cover trigger, discovery, syntax, thresholds, paths, resilience, mental model, recovery, completion and budget. \"Nothing found in this dimension\" is a valid answer; a list padded to a target length invites invented findings.\n5. **No errata files.** The corrected source is the deliverable. A line marked `[NICE]` is load-bearing: never weaken or delete it without a finding that supersedes it.\n\n## Every finding names two things\n\n- the **controlling text**: the exact sentence, example, row, help line or message that led the agent there - or its **absence** (`ABSENT`) when nothing said what was needed;\n- what the agent **believed** because of it, and whether that belief was `observed` (the trace shows reasoning) or `inferred` (reconstructed from what it did). Thinking text is usually not exposed, so most beliefs are inferred; say so.\n\nSchema and handback rules: [finding-schema](references\u002Ffinding-schema.md).\n\n\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#Severity & Root-Cause Cheat Sheet -->\n## Markers, severity, root cause (cheat sheet)\n\n| Marker | Severity | Typical root cause | Fix family |\n|---|---|---|---|\n| `[STUCK]` cannot continue | P0 | S1 missing prerequisite, S2 contradiction, M2 unstated location | Prerequisite Surfacing, Path Reconciliation, Output Location |\n| `[BROKE]` command or selector threw | P0 \u002F P1 | O1 silent crash, O5 stale flag, O3 unhandled edge case | Error Recovery Addition, Option Modernization |\n| `[GUESSED]` invented an unstated decision | P1 | M1 ambiguous threshold, M5 assumed knowledge | Threshold Concretization, Scaling Guidance |\n| Premature completion, skipped gates | P1 | C1 completion illusion, C4 gate slack | Two-Tier Verification Enclosure, Rigid Phase Gating |\n| Skipped destructive cleanup or mutation | P1 | C2 mutation hesitation | Staged Deprecation and Safe Neutralization |\n| Ran `--help` instead of a real probe | P1 | C3 missing tooling awareness, M4 missing method | Actionable Pre-Flight Probe |\n| Path drift or file recreation | P1 | S3 scattered information, M3 format inconsistency | Canonical Layout Enforcement |\n| Target never chosen, or chosen wrongly | P1 | T1 not triggered, T2 collision, T3 over-triggered | Description rewrite at the trigger layer |\n| `[NICE]` the text prevented an error | Keep | a load-bearing sentence or check | **Never weaken or delete** |\n\nThree or more P1 findings in one phase collapse into a **compound P0**. The full code list (S1-S5, M1-M7, O1-O8, C1-C5, T1-T3) is in [taxonomy](references\u002Ftaxonomy.md); the matching fix for each code is in [root-cause-matrix](references\u002Froot-cause-matrix.md) and [fix-patterns](references\u002Ffix-patterns.md).\n\n## Read when\n\n| File | Read when |\n|---|---|\n| [taxonomy](references\u002Ftaxonomy.md) | assigning a marker, severity and root-cause code |\n| [root-cause-matrix](references\u002Froot-cause-matrix.md) | choosing the fix for a code |\n| [fix-patterns](references\u002Ffix-patterns.md) | writing the rewrite itself |\n| [reading-traces](references\u002Freading-traces.md) | reading a trace: what drift looks like and where to look |\n| [evidence-and-anti-sycophancy](references\u002Fevidence-and-anti-sycophancy.md) | deciding whether \"done\" is true |\n| [finding-schema](references\u002Ffinding-schema.md) | producing or checking findings |\n","\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#Audit Skill by Derailment: The Native Two-Tier Auditor -->\n# Derailment method\n\nA skill, a tool description, a `--help` page and an error message are a program that an LLM executes. When the agent drifts, the defect is in the text or affordance it was given. This method finds that text from the evidence of a real run and rewrites it. Two tiers do the work: a **worker** (a fresh agent that attempts a realistic task) and a **reader** (a read-only agent that studies the worker's whole execution). The orchestrator applies fixes and reruns.\n\n\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#The 5 Non-Negotiable Invariants -->\n## The five invariants\n\n1. **Fix the text, not the executor.** Every remedy is an edit to the controlling text (a skill sentence, an example, a routing row, a tool description, an error message, a script) or to the tool itself. Never conclude \"use a smarter agent\". Rewrite or delete the sentence that misled; do not leave it and add a \"do \u002F don't\" rule beside it.\n2. **Claims are leads; the machine is the witness.** An agent's closing report is a claim. Verify exit codes and what is on disk (`agentic-maker check receipts`) before believing \"done\". See [evidence-and-anti-sycophancy](references\u002Fevidence-and-anti-sycophancy.md).\n3. **Never judge from a view that hides what the agent did.** Read the harness's structured record (every tool call with full arguments, every result, hook events). A folded terminal view, a summary or the agent's own account is not evidence. Source ranking: `skills\u002Fderail\u002Freferences\u002Ftrace-sources.md`.\n4. **Search by dimension; do not chase a quota.** Cover trigger, discovery, syntax, thresholds, paths, resilience, mental model, recovery, completion and budget. \"Nothing found in this dimension\" is a valid answer; a list padded to a target length invites invented findings.\n5. **No errata files.** The corrected source is the deliverable. A line marked `[NICE]` is load-bearing: never weaken or delete it without a finding that supersedes it.\n\n## Every finding names two things\n\n- the **controlling text**: the exact sentence, example, row, help line or message that led the agent there - or its **absence** (`ABSENT`) when nothing said what was needed;\n- what the agent **believed** because of it, and whether that belief was `observed` (the trace shows reasoning) or `inferred` (reconstructed from what it did). Thinking text is usually not exposed, so most beliefs are inferred; say so.\n\nSchema and handback rules: [finding-schema](references\u002Ffinding-schema.md).\n\n\u003C!-- src: audit-skill-by-derailment\u002FSKILL.md#Severity & Root-Cause Cheat Sheet -->\n## Markers, severity, root cause (cheat sheet)\n\n| Marker | Severity | Typical root cause | Fix family |\n|---|---|---|---|\n| `[STUCK]` cannot continue | P0 | S1 missing prerequisite, S2 contradiction, M2 unstated location | Prerequisite Surfacing, Path Reconciliation, Output Location |\n| `[BROKE]` command or selector threw | P0 \u002F P1 | O1 silent crash, O5 stale flag, O3 unhandled edge case | Error Recovery Addition, Option Modernization |\n| `[GUESSED]` invented an unstated decision | P1 | M1 ambiguous threshold, M5 assumed knowledge | Threshold Concretization, Scaling Guidance |\n| Premature completion, skipped gates | P1 | C1 completion illusion, C4 gate slack | Two-Tier Verification Enclosure, Rigid Phase Gating |\n| Skipped destructive cleanup or mutation | P1 | C2 mutation hesitation | Staged Deprecation and Safe Neutralization |\n| Ran `--help` instead of a real probe | P1 | C3 missing tooling awareness, M4 missing method | Actionable Pre-Flight Probe |\n| Path drift or file recreation | P1 | S3 scattered information, M3 format inconsistency | Canonical Layout Enforcement |\n| Target never chosen, or chosen wrongly | P1 | T1 not triggered, T2 collision, T3 over-triggered | Description rewrite at the trigger layer |\n| `[NICE]` the text prevented an error | Keep | a load-bearing sentence or check | **Never weaken or delete** |\n\nThree or more P1 findings in one phase collapse into a **compound P0**. The full code list (S1-S5, M1-M7, O1-O8, C1-C5, T1-T3) is in [taxonomy](references\u002Ftaxonomy.md); the matching fix for each code is in [root-cause-matrix](references\u002Froot-cause-matrix.md) and [fix-patterns](references\u002Ffix-patterns.md).\n\n## Read when\n\n| File | Read when |\n|---|---|\n| [taxonomy](references\u002Ftaxonomy.md) | assigning a marker, severity and root-cause code |\n| [root-cause-matrix](references\u002Froot-cause-matrix.md) | choosing the fix for a code |\n| [fix-patterns](references\u002Ffix-patterns.md) | writing the rewrite itself |\n| [reading-traces](references\u002Freading-traces.md) | reading a trace: what drift looks like and where to look |\n| [evidence-and-anti-sycophancy](references\u002Fevidence-and-anti-sycophancy.md) | deciding whether \"done\" is true |\n| [finding-schema](references\u002Ffinding-schema.md) | producing or checking findings |\n",{},"main","4fbc0dad2dfb64cb1747e69933c261d941c8f074","skills\u002Fderailment-method","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker","https:\u002F\u002Fgithub.com\u002Fyigitkonur","https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fyigitkonur\u002F","HKEY_LOCAL_MACHINE","2026-10-26 07:52:56",{"source":7,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add yigitkonur\u002Fplugin-agentic-maker --skill derailment-method",{"tool":37,"command":38},"bunx","bunx skills add yigitkonur\u002Fplugin-agentic-maker --skill derailment-method",{"tool":40,"command":41},"pnpm","pnpm dlx skills add yigitkonur\u002Fplugin-agentic-maker --skill derailment-method","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002FSKILL.md",true,[45,51,55,59,63,67],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"references\u002Fevidence-and-anti-sycophancy.md","REFERENCE","参考文档",4765,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Fevidence-and-anti-sycophancy.md",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Ffinding-schema.md",4086,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Ffinding-schema.md",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"references\u002Ffix-patterns.md",6593,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Ffix-patterns.md",{"path":60,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":61,"sourceHtmlUrl":62},"references\u002Freading-traces.md",5445,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Freading-traces.md",{"path":64,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":65,"sourceHtmlUrl":66},"references\u002Froot-cause-matrix.md",4591,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Froot-cause-matrix.md",{"path":68,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":69,"sourceHtmlUrl":70},"references\u002Ftaxonomy.md",7114,"https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderailment-method\u002Freferences\u002Ftaxonomy.md",[72,78,84,91,98,105],{"skillNo":73,"name":74,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":18},"01M7F6C40YCW0N1P9ENGE9WN50","design","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fdesign\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":18},"01M7F6BYP9843BZRKVG0K13HBK","targets","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderail\u002Freferences\u002Ftargets\u002Fskill.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":18},"01M7F6BTN9TZSN9QD6B5QG0JQD","derail","Test a skill, plugin, CLI or MCP tool with fresh agents and fix the text that misled them. Started by you; spends usage.","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fderail\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":18},"01M7F6BPR2E18652MMHFJZ7F15","build-skill","Create, redesign or merge a skill, grounded in facts, then hand over the test. Started by you.","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fbuild-skill\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":18},"01M7F6BK9QTJC0KR2GR6KGX4QQ","build-mcp","Build, extend or repair an MCP server for agents, probe it, hand over the test. Started by you.","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fbuild-mcp\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":18},"01M7F6BGP7DJJ3HJ604SF207TZ","build-cli","Build or repair an agent-facing CLI, probe it, hand over the test. Started by you.","https:\u002F\u002Fgithub.com\u002Fyigitkonur\u002Fplugin-agentic-maker\u002Fblob\u002Fmain\u002Fskills\u002Fbuild-cli\u002FSKILL.md",[],[]]