[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C9F93YCXFCVRDG4S5QFAH4":3,"skill-same:01M7C9F93YCXFCVRDG4S5QFAH4":58},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7C9F93YCXFCVRDG4S5QFAH4","skill-authoring-edho-ferdian",null,"Create and govern this ecosystem's skills: search before building, a quality bar, compliance checks, promoting repeated principles to rules, and packaging dist\u002F*.skill. Use for \"bikin skill baru\", \"ada skill buat X gak\", \"fork skill ini\", \"skill gue kepake gak sih\", \"package skill ini\", \"buatkan .skill-nya\".","edhoferdian\u002FEEF","edhoferdian","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F64404158?v=4","01M7C8538ZVVZWZEN1AW2MQXCV","Edho F","01M7C8539NS56BMJ4GXN8A47TQ",22,"MIT","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring-edho-ferdian\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: skill-authoring-edho-ferdian\ndescription: >-\n  Create and govern this ecosystem's skills: search before building, a\n  quality bar, compliance checks, promoting repeated principles to rules,\n  and packaging dist\u002F*.skill. Use for \"bikin skill baru\", \"ada skill buat X\n  gak\", \"fork skill ini\", \"skill gue kepake gak sih\", \"package skill ini\",\n  \"buatkan .skill-nya\".\nmetadata:\n  last_reviewed: \"2026-10-08\"\n---\n\n# Skill Authoring — Edho Ferdian Mode\n\n## Boundary with `skill-audit-edho-ferdian`\n\n| Moment | Skill |\n|---|---|\n| Before a skill exists — should it? does one already? | **this skill** §1 |\n| While writing it — quality bar, description that triggers | **this skill** §2 |\n| After it ships — is it actually obeyed? | **this skill** §3 |\n| Periodically across all skills — stale, redundant, broken links | `skill-audit-edho-ferdian` |\n| Principles recurring across many skills → rules | **this skill** §4 |\n| Packaging a finished skill into `dist\u002F*.skill` | **this skill** §6 |\n\nNeither skill runs the other's checks. If you find yourself grading\ndescription quality across the whole `skills\u002F` tree, you are running the\naudit, not this.\n\nAlso distinct from `opensource-release-edho-ferdian`: that skill packages\n*someone else's project* for public open-source release. §6 below packages\n*this ecosystem's own* skill folders for distribution. Same verb\n(\"package\"), different object — don't route a \"package this skill\" request\nthere.\n\n## §1 — Search before you build\n\nThe failure mode is building the twelfth variant of something that already\nexists. Search in cost order, cheapest first:\n\n1. **This ecosystem** — `skills\u002F*\u002FSKILL.md` in this repo. A near-match here\n   is usually a FOLD (a new `references\u002F*.md` under an existing skill), not\n   a new top-level skill. This is the same consolidation rule D-009 set for\n   this ecosystem's earlier porting work, applied to original work.\n2. **Installed and marketplace skills** — name match first, then\n   frontmatter descriptions.\n3. **GitHub** — `gh search repos`, `gh search code --filename SKILL.md`.\n4. **Web** — at most three targeted queries.\n\n**Vet anything external before adopting it.** Read the full SKILL.md and\nevery reference. Look for unexpected shell commands, file writes outside\nthe working tree, network calls, credential handling, or package installs.\nCheck whether the repo is maintained. Copy into a fresh branch and review\nthe diff rather than editing a marketplace original in place. An external\nskill is untrusted content until read — the same standard this ecosystem\nholds for any other fetched file.\n\nIf the user explicitly says to skip the search, acknowledge it and proceed.\n\n**Choosing the right artifact form.** A repeated pattern becomes a\n**command** when a human decides to invoke it at a known moment; a **skill**\nwhen it should fire on its own from a description-level trigger; an\n**agent\u002Fsubagent** when it needs its own context window or an isolated\nwrite surface. Choosing \"skill\" for something that is really an on-demand\nchecklist is what produces trigger collisions later — see `skill-audit-\nedho-ferdian`'s redundancy category.\n\n## §2 — The quality bar\n\nA skill in this ecosystem is not done until:\n\n- **The description triggers.** It names concrete phrases the user actually\n  says (including Indonesian ones), and it says what the skill is *not* for\n  when a neighbouring skill exists. A description that is merely accurate\n  but never fires is a dead skill.\n- **The description is short.** At most 450 characters (CI enforces it via\n  `measure_context.py --max-desc-chars`): what the skill does, then the\n  trigger phrases. Claude Code caps the whole skill listing and drops the\n  descriptions of rarely used skills when it overflows, so a long\n  description can cost another skill its triggers. Boundary detail goes in\n  the body.\n- **Its boundary is stated.** If it overlaps another skill, the overlap is\n  named and adjudicated in the text — the audit skill treats a stated\n  boundary as intentional architecture, not redundancy.\n- **Provenance is recorded** when adapted from an external source: what it\n  was adapted from, the fetch date, and what was deliberately changed.\n- **References are split by lens, not by chapter.** Many small\n  `references\u002F*.md` loaded on demand beats one long SKILL.md.\n- **No live dependency on an external harness** (D-005), and no\n  Claude-Code-only construct that breaks portability (D-008).\n\n## §3 — Compliance: does anyone actually obey it?\n\nThe unexamined assumption behind every skill collection is that written\ninstructions are followed. They frequently are not, and nobody notices because nobody looks.\n\nTest one skill by running the same task at three prompt strictness levels\nand classifying what the agent actually did:\n\n| Level | Prompt shape | What a pass looks like |\n|---|---|---|\n| Supportive | Names the skill explicitly | The skill's steps appear in order |\n| Neutral | Describes the task using the skill's own trigger phrases, without naming it | The skill still fires |\n| Competing | Describes the task while nudging toward a shortcut the skill forbids | The skill's constraint holds |\n\nThe neutral level tests the description; the competing level tests whether\nthe rules are load-bearing or decorative. Report the compliance rate and\nthe tool-call sequence, not an impression. A skill that only passes at the\nsupportive level has a description problem; one that fails at the competing\nlevel has a rules problem.\n\n## §4 — Distilling rules from skills\n\nWhen the same principle appears in three or more skills, it belongs in a\nrule, not repeated in each. This is the same promotion rule Stage 6 of\n`dev-kickoff-edho-ferdian` applies to instincts (\"three or more related\ninstincts pointing the same way → propose promoting them into PDR §3\"),\napplied one level up.\n\nMethod: collect exhaustively and mechanically (grep for the repeated\nphrasing across all `skills\u002F**\u002F*.md`), then judge with full context —\nscripts gather facts, the model decides. Output is one of: append to an\nexisting rule file, revise an outdated one, or create a new one. Replace\nthe now-duplicated passages with a cross-reference rather than leaving both.\n\n## §5 — Incremental audit handoff\n\nA full pass over every skill is\nexpensive and mostly re-reads unchanged files. Prefer a **quick scan**:\ncompare each `SKILL.md` and `references\u002F*.md` mtime against the last audit\nrun, re-evaluate only what changed, and carry forward previous findings for\nthe rest. Report the diff. Fall back to a full pass when no previous run\nexists, or on explicit request.\n\nThe findings format and the four audit categories belong to\n`skill-audit-edho-ferdian` — this section only governs *which files* that\naudit needs to look at.\n\n## §6 — Packaging for distribution (`dist\u002F*.skill`)\n\nPackages *this ecosystem's own* skill folders for manual upload (Claude.ai \u002F\nClaude Desktop \u002F Claude Code Skills UI) — not `opensource-release-edho-\nferdian`'s job. Whenever `skills\u002F\u003Cname>\u002F` changes and is about to be\ncommitted: run `python scripts\u002Fvalidate_skills.py`, then\n`python scripts\u002Fpackage_skills.py` (all skills, or `\u003Cname>` for one;\n`--check` is the dry run CI's `dist-sync` job uses). Never hand-zip a skill\nfolder. Publishing the archive anywhere is a separate, explicit-permission\naction. What the validator does and doesn't check, and why the script is\nthe only packaging path: **`references\u002Fpackaging.md`**.\n\n## §7 — Language routing (canonical contract — all skills point here)\n\nThe single source of truth for which human language a *shipped, installed*\nskill uses wherever it runs. It is not this repo's own `CLAUDE.md` §F (which\ngoverns curation work inside this repo) — the two agree in value but are\nseparate documents; don't merge or delete either. Why this was promoted and\nthe full scope note: **`references\u002Fcontract-history.md`**.\n\n**The base rule (fixed, never ask):**\n1. Narration, explanations, questions, and reports to the user → **Bahasa\n   Indonesia**.\n2. Code, diffs, commit messages, filenames, folder names, and any other\n   machine-facing generated artifact → **English**.\n3. Never ask the user which language to use — this is fixed, not a\n   preference to elicit.\n\n**When a skill's own source material carries a language** (e.g. a spec,\ndocument, or dataset already written in a specific language it must mirror\nback), detect that source language and follow it for content that mirrors\nthe source, while the base rule above still governs narration and\nartifacts. `dev-kickoff-edho-ferdian`'s \"Language routing\" section is the\nfullest worked example of this split (`doc_lang` vs `artifact_lang`,\nbecause it ingests specs that may already be Indonesian or English) — read\nit before writing a new multi-language exception rather than re-deriving\none from scratch.\n\n**Standard form for every other skill** (one heading, 1-2 sentences, no\ninline restatement of the full contract):\n`## Language routing (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming the base rule and pointing here.\n\n## §8 — Development loop convention (canonical contract — orchestrating skills point here)\n\nEvery skill in this ecosystem that drives multi-step implementation work —\nnot a single-purpose lens like `code-review-edho-ferdian`, but a skill that\nplans, writes, and closes out a task — follows the same seven-stage loop:\n\n```\nPLAN → TEST → IMPLEMENT → REVIEW → VERIFY → REMEMBER → IMPROVE\n```\n\n`dev-kickoff-edho-ferdian` is the canonical, fully-specified implementation\nof this loop (`references\u002Fexecution-loop.md` there has the complete\nper-stage protocol, gate list, and the auto-invocation contract that wires\neach stage to the matching specialist skill in this ecosystem — read it\nbefore re-deriving any part of this cycle elsewhere). Two rules apply to any\nother skill that adopts this loop:\n\n1. **Point, don't restate.** State the loop in one line and point to\n   `dev-kickoff-edho-ferdian`'s `references\u002Fexecution-loop.md`, the same way\n   every skill states Language routing in one line and points to §7. A\n   second full copy of the per-stage protocol is a duplicate that will drift.\n2. **The last two stages are not optional decoration.** REMEMBER without\n   IMPROVE records a lesson that never changes anything; a skill that\n   implements PLAN through REMEMBER but drops IMPROVE has built a diary, not\n   a feedback loop. If a skill's scope genuinely has no instincts\u002Fdebt\u002F\n   dead-code surface to check (rare — most implementation work does), say so\n   explicitly rather than silently truncating the cycle to six stages.\n\n**Auto-invocation is the same idea as the \"search before you build\" (§1)\ndiscipline, applied mid-task instead of pre-build.** A skill mid-loop that\nhits a stage matching another skill's specialty (writing a test → consult\n`test-authoring-edho-ferdian`; reviewing code → consult\n`code-review-edho-ferdian`; a build breaks → consult `build-fix-edho-ferdian`)\ninvokes that skill rather than reproducing its judgment from general\nknowledge. This is what makes the ecosystem behave as one system instead of\n33 isolated documents that happen to share a naming suffix.\n\n## §9 — External docs lookup (Context7) (canonical contract — authoring skills point here)\n\nContext7 (`mcp__context7__resolve-library-id` → `query-docs`) is a third live\nexternal dependency for this ecosystem, alongside ECC (D-005) and Salak\n(D-004) — consumed, never rebuilt (D-044). It fetches current documentation\nfor a library\u002Fframework\u002FSDK\u002FAPI\u002FCLI directly from source, correcting for\ntraining-data staleness on version-specific API surfaces.\n\n**When to call it (authoring-time only, not every line of code):**\n1. Adding a new dependency or integration.\n2. Upgrading or migrating a dependency's major\u002Fminor version.\n3. Using an API surface that changes fast, or is used rarely enough that a\n   memorized signature is unreliable.\n4. Diagnosing a build\u002Fcompile error that is actually a signature or\n   version-drift issue (`build-fix-edho-ferdian`'s narrow use case).\n\nDo not call it for refactoring, business-logic debugging, code review, or\ngeneral programming concepts that don't hinge on a specific library's\ncurrent API — this matches the boundary the Context7 MCP server itself\ndeclares.\n\n**Session discipline:**\n- Resolve a library's ID once per session and reuse it across every\n  `query-docs` call for that library — don't re-resolve on every question.\n- On rate-limit or failure: fall back to `WebFetch` against the library's\n  official docs directly; if that also fails, answer from training\n  knowledge but say explicitly it is unverified and may be stale. Never\n  silently answer as if it were confirmed current.\n- **No local caching of fetched docs** — a local copy quietly goes stale,\n  the same burden this ecosystem rejected for ECC and Salak. Query live.\n\n**Rate limits.** The anonymous\u002Fno-key connection is rate-limited; a free\nContext7 API key raises the limit. Adding one is a manual user action — the\nexecutor never stores a key or reconfigures the MCP server unattended.\nSetup steps: **`references\u002Fcontract-history.md`**.\n\n**Standard form for every other skill** (one heading, one sentence, pointing\nhere — same convention as §7\u002F§8):\n`## External docs (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming when this skill calls Context7 and\npointing here.\n\n## §10 — Surgical changes (canonical contract — any code-editing skill points here)\n\nA general default for editing existing code, distinct from any single\nskill's own workflow: **touch only what the user's current instruction\nrequires.** This governs every skill that edits code, not just the ones\nthat already name it below.\n\n**The rule:**\n1. Don't \"improve\" adjacent code, comments, or formatting while you're in\n   a file for an unrelated reason. Match existing style even where you'd\n   choose differently — that choice belongs to a separate, explicit task.\n2. Don't refactor something that isn't broken as a side effect of fixing\n   or extending something else.\n3. When your own change makes an import, variable, or function unused,\n   remove it — but don't remove *pre-existing* dead code you merely\n   noticed. Flag it (or hand off to `dead-code-cleanup-edho-ferdian`)\n   instead of pulling it into scope.\n4. The test that decides an edge case: every changed line should trace\n   directly to the user's current request. If it doesn't, it's a separate\n   task — mention it, don't fold it in silently.\n\n**This is a default, not an absolute** — a skill with its own stricter\nversion keeps it (e.g. `build-fix-edho-ferdian`'s Phase 3 \"never touch\nunrelated code, even a one-line obvious improvement\" sharpens it for the\nbuild-fix moment). Skills that flag violations at review time\n(`code-review-edho-ferdian`'s BC-08 \"scope creep\", `dev-kickoff-edho-\nferdian`'s \"silent scope creep\" anti-pattern) keep doing so — this section\nis the authoring-time default they catch violations of.\n\n**Standard form for every other skill** (one heading, one sentence, pointing\nhere — same convention as §7\u002F§8\u002F§9):\n`## Surgical changes (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming how this skill's own scope\ndiscipline relates to the default (inherits it as-is, or sharpens it for\nthis skill's specific moment, per the build-fix example above).\n\n## Provenance\n\nConsolidated into one skill per D-009; install-specific paths were replaced\nwith this repo's own `skills\u002F` tree. When and why each of §6–§10 was added,\nincluding §10's external source and what was deliberately not ported:\n**`references\u002Fcontract-history.md`**.\n","# Skill Authoring — Edho Ferdian Mode\n\n## Boundary with `skill-audit-edho-ferdian`\n\n| Moment | Skill |\n|---|---|\n| Before a skill exists — should it? does one already? | **this skill** §1 |\n| While writing it — quality bar, description that triggers | **this skill** §2 |\n| After it ships — is it actually obeyed? | **this skill** §3 |\n| Periodically across all skills — stale, redundant, broken links | `skill-audit-edho-ferdian` |\n| Principles recurring across many skills → rules | **this skill** §4 |\n| Packaging a finished skill into `dist\u002F*.skill` | **this skill** §6 |\n\nNeither skill runs the other's checks. If you find yourself grading\ndescription quality across the whole `skills\u002F` tree, you are running the\naudit, not this.\n\nAlso distinct from `opensource-release-edho-ferdian`: that skill packages\n*someone else's project* for public open-source release. §6 below packages\n*this ecosystem's own* skill folders for distribution. Same verb\n(\"package\"), different object — don't route a \"package this skill\" request\nthere.\n\n## §1 — Search before you build\n\nThe failure mode is building the twelfth variant of something that already\nexists. Search in cost order, cheapest first:\n\n1. **This ecosystem** — `skills\u002F*\u002FSKILL.md` in this repo. A near-match here\n   is usually a FOLD (a new `references\u002F*.md` under an existing skill), not\n   a new top-level skill. This is the same consolidation rule D-009 set for\n   this ecosystem's earlier porting work, applied to original work.\n2. **Installed and marketplace skills** — name match first, then\n   frontmatter descriptions.\n3. **GitHub** — `gh search repos`, `gh search code --filename SKILL.md`.\n4. **Web** — at most three targeted queries.\n\n**Vet anything external before adopting it.** Read the full SKILL.md and\nevery reference. Look for unexpected shell commands, file writes outside\nthe working tree, network calls, credential handling, or package installs.\nCheck whether the repo is maintained. Copy into a fresh branch and review\nthe diff rather than editing a marketplace original in place. An external\nskill is untrusted content until read — the same standard this ecosystem\nholds for any other fetched file.\n\nIf the user explicitly says to skip the search, acknowledge it and proceed.\n\n**Choosing the right artifact form.** A repeated pattern becomes a\n**command** when a human decides to invoke it at a known moment; a **skill**\nwhen it should fire on its own from a description-level trigger; an\n**agent\u002Fsubagent** when it needs its own context window or an isolated\nwrite surface. Choosing \"skill\" for something that is really an on-demand\nchecklist is what produces trigger collisions later — see `skill-audit-\nedho-ferdian`'s redundancy category.\n\n## §2 — The quality bar\n\nA skill in this ecosystem is not done until:\n\n- **The description triggers.** It names concrete phrases the user actually\n  says (including Indonesian ones), and it says what the skill is *not* for\n  when a neighbouring skill exists. A description that is merely accurate\n  but never fires is a dead skill.\n- **The description is short.** At most 450 characters (CI enforces it via\n  `measure_context.py --max-desc-chars`): what the skill does, then the\n  trigger phrases. Claude Code caps the whole skill listing and drops the\n  descriptions of rarely used skills when it overflows, so a long\n  description can cost another skill its triggers. Boundary detail goes in\n  the body.\n- **Its boundary is stated.** If it overlaps another skill, the overlap is\n  named and adjudicated in the text — the audit skill treats a stated\n  boundary as intentional architecture, not redundancy.\n- **Provenance is recorded** when adapted from an external source: what it\n  was adapted from, the fetch date, and what was deliberately changed.\n- **References are split by lens, not by chapter.** Many small\n  `references\u002F*.md` loaded on demand beats one long SKILL.md.\n- **No live dependency on an external harness** (D-005), and no\n  Claude-Code-only construct that breaks portability (D-008).\n\n## §3 — Compliance: does anyone actually obey it?\n\nThe unexamined assumption behind every skill collection is that written\ninstructions are followed. They frequently are not, and nobody notices because nobody looks.\n\nTest one skill by running the same task at three prompt strictness levels\nand classifying what the agent actually did:\n\n| Level | Prompt shape | What a pass looks like |\n|---|---|---|\n| Supportive | Names the skill explicitly | The skill's steps appear in order |\n| Neutral | Describes the task using the skill's own trigger phrases, without naming it | The skill still fires |\n| Competing | Describes the task while nudging toward a shortcut the skill forbids | The skill's constraint holds |\n\nThe neutral level tests the description; the competing level tests whether\nthe rules are load-bearing or decorative. Report the compliance rate and\nthe tool-call sequence, not an impression. A skill that only passes at the\nsupportive level has a description problem; one that fails at the competing\nlevel has a rules problem.\n\n## §4 — Distilling rules from skills\n\nWhen the same principle appears in three or more skills, it belongs in a\nrule, not repeated in each. This is the same promotion rule Stage 6 of\n`dev-kickoff-edho-ferdian` applies to instincts (\"three or more related\ninstincts pointing the same way → propose promoting them into PDR §3\"),\napplied one level up.\n\nMethod: collect exhaustively and mechanically (grep for the repeated\nphrasing across all `skills\u002F**\u002F*.md`), then judge with full context —\nscripts gather facts, the model decides. Output is one of: append to an\nexisting rule file, revise an outdated one, or create a new one. Replace\nthe now-duplicated passages with a cross-reference rather than leaving both.\n\n## §5 — Incremental audit handoff\n\nA full pass over every skill is\nexpensive and mostly re-reads unchanged files. Prefer a **quick scan**:\ncompare each `SKILL.md` and `references\u002F*.md` mtime against the last audit\nrun, re-evaluate only what changed, and carry forward previous findings for\nthe rest. Report the diff. Fall back to a full pass when no previous run\nexists, or on explicit request.\n\nThe findings format and the four audit categories belong to\n`skill-audit-edho-ferdian` — this section only governs *which files* that\naudit needs to look at.\n\n## §6 — Packaging for distribution (`dist\u002F*.skill`)\n\nPackages *this ecosystem's own* skill folders for manual upload (Claude.ai \u002F\nClaude Desktop \u002F Claude Code Skills UI) — not `opensource-release-edho-\nferdian`'s job. Whenever `skills\u002F\u003Cname>\u002F` changes and is about to be\ncommitted: run `python scripts\u002Fvalidate_skills.py`, then\n`python scripts\u002Fpackage_skills.py` (all skills, or `\u003Cname>` for one;\n`--check` is the dry run CI's `dist-sync` job uses). Never hand-zip a skill\nfolder. Publishing the archive anywhere is a separate, explicit-permission\naction. What the validator does and doesn't check, and why the script is\nthe only packaging path: **`references\u002Fpackaging.md`**.\n\n## §7 — Language routing (canonical contract — all skills point here)\n\nThe single source of truth for which human language a *shipped, installed*\nskill uses wherever it runs. It is not this repo's own `CLAUDE.md` §F (which\ngoverns curation work inside this repo) — the two agree in value but are\nseparate documents; don't merge or delete either. Why this was promoted and\nthe full scope note: **`references\u002Fcontract-history.md`**.\n\n**The base rule (fixed, never ask):**\n1. Narration, explanations, questions, and reports to the user → **Bahasa\n   Indonesia**.\n2. Code, diffs, commit messages, filenames, folder names, and any other\n   machine-facing generated artifact → **English**.\n3. Never ask the user which language to use — this is fixed, not a\n   preference to elicit.\n\n**When a skill's own source material carries a language** (e.g. a spec,\ndocument, or dataset already written in a specific language it must mirror\nback), detect that source language and follow it for content that mirrors\nthe source, while the base rule above still governs narration and\nartifacts. `dev-kickoff-edho-ferdian`'s \"Language routing\" section is the\nfullest worked example of this split (`doc_lang` vs `artifact_lang`,\nbecause it ingests specs that may already be Indonesian or English) — read\nit before writing a new multi-language exception rather than re-deriving\none from scratch.\n\n**Standard form for every other skill** (one heading, 1-2 sentences, no\ninline restatement of the full contract):\n`## Language routing (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming the base rule and pointing here.\n\n## §8 — Development loop convention (canonical contract — orchestrating skills point here)\n\nEvery skill in this ecosystem that drives multi-step implementation work —\nnot a single-purpose lens like `code-review-edho-ferdian`, but a skill that\nplans, writes, and closes out a task — follows the same seven-stage loop:\n\n```\nPLAN → TEST → IMPLEMENT → REVIEW → VERIFY → REMEMBER → IMPROVE\n```\n\n`dev-kickoff-edho-ferdian` is the canonical, fully-specified implementation\nof this loop (`references\u002Fexecution-loop.md` there has the complete\nper-stage protocol, gate list, and the auto-invocation contract that wires\neach stage to the matching specialist skill in this ecosystem — read it\nbefore re-deriving any part of this cycle elsewhere). Two rules apply to any\nother skill that adopts this loop:\n\n1. **Point, don't restate.** State the loop in one line and point to\n   `dev-kickoff-edho-ferdian`'s `references\u002Fexecution-loop.md`, the same way\n   every skill states Language routing in one line and points to §7. A\n   second full copy of the per-stage protocol is a duplicate that will drift.\n2. **The last two stages are not optional decoration.** REMEMBER without\n   IMPROVE records a lesson that never changes anything; a skill that\n   implements PLAN through REMEMBER but drops IMPROVE has built a diary, not\n   a feedback loop. If a skill's scope genuinely has no instincts\u002Fdebt\u002F\n   dead-code surface to check (rare — most implementation work does), say so\n   explicitly rather than silently truncating the cycle to six stages.\n\n**Auto-invocation is the same idea as the \"search before you build\" (§1)\ndiscipline, applied mid-task instead of pre-build.** A skill mid-loop that\nhits a stage matching another skill's specialty (writing a test → consult\n`test-authoring-edho-ferdian`; reviewing code → consult\n`code-review-edho-ferdian`; a build breaks → consult `build-fix-edho-ferdian`)\ninvokes that skill rather than reproducing its judgment from general\nknowledge. This is what makes the ecosystem behave as one system instead of\n33 isolated documents that happen to share a naming suffix.\n\n## §9 — External docs lookup (Context7) (canonical contract — authoring skills point here)\n\nContext7 (`mcp__context7__resolve-library-id` → `query-docs`) is a third live\nexternal dependency for this ecosystem, alongside ECC (D-005) and Salak\n(D-004) — consumed, never rebuilt (D-044). It fetches current documentation\nfor a library\u002Fframework\u002FSDK\u002FAPI\u002FCLI directly from source, correcting for\ntraining-data staleness on version-specific API surfaces.\n\n**When to call it (authoring-time only, not every line of code):**\n1. Adding a new dependency or integration.\n2. Upgrading or migrating a dependency's major\u002Fminor version.\n3. Using an API surface that changes fast, or is used rarely enough that a\n   memorized signature is unreliable.\n4. Diagnosing a build\u002Fcompile error that is actually a signature or\n   version-drift issue (`build-fix-edho-ferdian`'s narrow use case).\n\nDo not call it for refactoring, business-logic debugging, code review, or\ngeneral programming concepts that don't hinge on a specific library's\ncurrent API — this matches the boundary the Context7 MCP server itself\ndeclares.\n\n**Session discipline:**\n- Resolve a library's ID once per session and reuse it across every\n  `query-docs` call for that library — don't re-resolve on every question.\n- On rate-limit or failure: fall back to `WebFetch` against the library's\n  official docs directly; if that also fails, answer from training\n  knowledge but say explicitly it is unverified and may be stale. Never\n  silently answer as if it were confirmed current.\n- **No local caching of fetched docs** — a local copy quietly goes stale,\n  the same burden this ecosystem rejected for ECC and Salak. Query live.\n\n**Rate limits.** The anonymous\u002Fno-key connection is rate-limited; a free\nContext7 API key raises the limit. Adding one is a manual user action — the\nexecutor never stores a key or reconfigures the MCP server unattended.\nSetup steps: **`references\u002Fcontract-history.md`**.\n\n**Standard form for every other skill** (one heading, one sentence, pointing\nhere — same convention as §7\u002F§8):\n`## External docs (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming when this skill calls Context7 and\npointing here.\n\n## §10 — Surgical changes (canonical contract — any code-editing skill points here)\n\nA general default for editing existing code, distinct from any single\nskill's own workflow: **touch only what the user's current instruction\nrequires.** This governs every skill that edits code, not just the ones\nthat already name it below.\n\n**The rule:**\n1. Don't \"improve\" adjacent code, comments, or formatting while you're in\n   a file for an unrelated reason. Match existing style even where you'd\n   choose differently — that choice belongs to a separate, explicit task.\n2. Don't refactor something that isn't broken as a side effect of fixing\n   or extending something else.\n3. When your own change makes an import, variable, or function unused,\n   remove it — but don't remove *pre-existing* dead code you merely\n   noticed. Flag it (or hand off to `dead-code-cleanup-edho-ferdian`)\n   instead of pulling it into scope.\n4. The test that decides an edge case: every changed line should trace\n   directly to the user's current request. If it doesn't, it's a separate\n   task — mention it, don't fold it in silently.\n\n**This is a default, not an absolute** — a skill with its own stricter\nversion keeps it (e.g. `build-fix-edho-ferdian`'s Phase 3 \"never touch\nunrelated code, even a one-line obvious improvement\" sharpens it for the\nbuild-fix moment). Skills that flag violations at review time\n(`code-review-edho-ferdian`'s BC-08 \"scope creep\", `dev-kickoff-edho-\nferdian`'s \"silent scope creep\" anti-pattern) keep doing so — this section\nis the authoring-time default they catch violations of.\n\n**Standard form for every other skill** (one heading, one sentence, pointing\nhere — same convention as §7\u002F§8\u002F§9):\n`## Surgical changes (fixed — see skill-authoring-edho-ferdian's canonical\ncontract)` followed by a sentence naming how this skill's own scope\ndiscipline relates to the default (inherits it as-is, or sharpens it for\nthis skill's specific moment, per the build-fix example above).\n\n## Provenance\n\nConsolidated into one skill per D-009; install-specific paths were replaced\nwith this repo's own `skills\u002F` tree. When and why each of §6–§10 was added,\nincluding §10's external source and what was deliberately not ported:\n**`references\u002Fcontract-history.md`**.\n",{"name":5,"metadata":24,"description":7},{"last_reviewed":25},"2026-10-08","main","533d32ad55c92d40fbabe93cbe21fe048598fe5a","skills\u002Fskill-authoring-edho-ferdian",0,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF","https:\u002F\u002Fgithub.com\u002Fedhoferdian","Elevating ideas through AI, rising from humble beginnings to mastering the frontiers of the web. Every line of code is open-sourced on GitHub, inviting you to f","2026-10-30 03:34:17",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add edhoferdian\u002FEEF --skill skill-authoring-edho-ferdian",{"tool":40,"command":41},"bunx","bunx skills add edhoferdian\u002FEEF --skill skill-authoring-edho-ferdian",{"tool":43,"command":44},"pnpm","pnpm dlx skills add edhoferdian\u002FEEF --skill skill-authoring-edho-ferdian","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring-edho-ferdian\u002FSKILL.md",true,[48,54],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fcontract-history.md","REFERENCE","参考文档",4920,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring-edho-ferdian\u002Freferences\u002Fcontract-history.md",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fpackaging.md",3018,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring-edho-ferdian\u002Freferences\u002Fpackaging.md",[59,66,73,80,87,94],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M7C9FY2EWE5NB7M3YW0KJX62","video-style-edho-ferdian","Measure a reference video's look and pacing (colour grade as a 3D LUT, cut rhythm, overlays) and reproduce it: grade, cut and composite new footage, verify numerically, hand off an NLE timeline. Use for \"tiru look video ini\", \"samakan grading\", \"bikin video dengan gaya referensi\", \"LUT dari referensi\", \"cut rhythm\".","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fvideo-style-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M7C9FT74P7JXGQ73116B9WS1","test-authoring-edho-ferdian","Write unit and component tests well: React Testing Library, pytest, Go and Vue, plus regression-test patterns. Use for \"tulis test untuk component ini\", \"bagaimana test hook ini\", \"test yang bagus untuk fitur X\", \"tulis test pytest\u002FGo\u002FVue\", or dev-kickoff's TEST stage.","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Ftest-authoring-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":19},"01M7C9FMPE9W76FS7S8ZZ48ARE","system-design-edho-ferdian","Mid-project architecture decisions: ADRs, trade-off analysis, non-functional requirements, scaling tiers and blast radius. Use for \"desain arsitektur\", \"keputusan teknis besar\", \"bikin ADR\", \"trade-off antara X dan Y\", \"microservices atau monolith\", \"apa yang rusak kalau X diubah\", or a capacity question.","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fsystem-design-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M7C9FETATAHKG5WKPT4AP1A2","spec-mining-edho-ferdian","Extract behavioural specs from a codebase with none: WHEN\u002FTHEN requirements and invariants, each anchored to the code that enforces it, unclear behaviour recorded as uncertainty. Use for \"ekstrak spec\", \"buat spec dari kode\", \"reverse-engineer the spec\", \"repo ini tidak ada dokumentasi\", or a missing BEHAVIOR_SPEC in dev-kickoff.","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fspec-mining-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":19},"01M7C9F4XZJ59R0GS4YEP4E1RD","skill-audit-edho-ferdian","Audit this repo's skills\u002F for staleness, redundancy, broken cross-references and weak descriptions (not the ~\u002F.claude setup: that is config-hygiene). Use for \"audit skill saya\", \"cek skill yang sudah dibuat\", \"ada yang redundan gak\", \"skill mana yang basi\".","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-audit-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M7C9F058P2C8PMN6ZH5EX8QV","seo-audit-edho-ferdian","Technical and on-page SEO audit: crawlability, indexability, structured data, meta tags, sitemap\u002Frobots.txt, mobile and internal links, ranked by indexing impact. Use for \"audit SEO\", \"kenapa website ini tidak muncul di Google\", \"cek meta tags\", \"structured data\", \"sitemap\u002Frobots.txt\", \"cek SEO\", including from pasted HTML.","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fseo-audit-edho-ferdian\u002FSKILL.md",[],[]]