[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C9BAV7VSFFBQW68ZDFZ1X7":3,"skill-same:01M7C9BAV7VSFFBQW68ZDFZ1X7":68},{"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},"01M7C9BAV7VSFFBQW68ZDFZ1X7","config-hygiene-edho-ferdian",null,"Clean up the ~\u002F.claude environment: redundant, stale or context-expensive skills, memory, hooks, permissions, MCP servers and caches, one item at a time with an undo path. Use for \"bersihin config\", \"~\u002F.claude berantakan\", \"kebanyakan skill\", \"sesi lambat mulai\", \"context cepat penuh\", \"audit setup gue\".","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\u002Fconfig-hygiene-edho-ferdian\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: config-hygiene-edho-ferdian\ndescription: >-\n  Clean up the ~\u002F.claude environment: redundant, stale or context-expensive\n  skills, memory, hooks, permissions, MCP servers and caches, one item at a\n  time with an undo path. Use for \"bersihin config\", \"~\u002F.claude berantakan\",\n  \"kebanyakan skill\", \"sesi lambat mulai\", \"context cepat penuh\", \"audit\n  setup gue\".\nmetadata:\n  last_reviewed: \"2026-10-27\"\n---\n\n# Config Hygiene — Edho Ferdian Mode\n\n## Scope — read this first\n\n**Target: the user's own Claude Code environment** — `~\u002F.claude\u002F*`,\nproject-level `.claude\u002F*`, `~\u002F.claude.json`, and `.mcp.json`.\n\nThis is the deliberate complement to `skill-audit-edho-ferdian`, which\naudits the *content quality* of this ecosystem's `skills\u002F*` and explicitly\nrefuses to look at `~\u002F.claude`. This skill looks at exactly what that one\nrefuses to: the environment those skills run inside.\n\n| Question | Skill |\n|---|---|\n| Is this ecosystem's skill well-written, non-redundant, links intact? | `skill-audit-edho-ferdian` |\n| Should a new skill exist at all, and is it actually obeyed? | `skill-authoring-edho-ferdian` |\n| Is the environment carrying dead weight, and what does it cost? | **this skill** |\n\nNever delete autonomously. Every candidate gets its own confirmation.\n\n## Design principles\n\n1. **Append-only configs leak.** Skills, memory files, hooks and permission\n   entries only ever get added. Without a periodic pass they rot silently.\n2. **Periodic beats one-time.** Scan roughly every 30 days, propose a small\n   batch each time. Cap a run at ~20 candidates — this is GC, not a purge.\n3. **Per-channel signals.** Each accumulation type has its own staleness\n   test. Do not apply one rule everywhere.\n4. **Soft-delete first.** Rename to `.disabled` > move to\n   `~\u002F.claude\u002F_gc_trash\u002F\u003Cdate>\u002F` > real deletion. Always keep an undo path.\n5. **Human-in-the-loop, per item.** Show the evidence, then `[y\u002Fn\u002Fskip]`.\n   There is no \"yes to all\".\n6. **Log every run** to `~\u002F.claude\u002Fgc_log.md`: what was touched, why, and\n   how to undo it.\n\n## Scan channels\n\n| # | Channel | Path | Staleness \u002F redundancy signals |\n|---|---|---|---|\n| 1 | Skills | `~\u002F.claude\u002Fskills\u002F*\u002F` | Heavily overlapping names; never triggered in recent transcripts; domain mismatch with actual work; broken or empty SKILL.md |\n| 2 | Memory | `~\u002F.claude\u002F**\u002Fmemory\u002F*.md` + index | Several index entries for one topic; contents contradicted by newer entries; dates already passed; orphans missing from the index; sub-100-word fragments that should merge |\n| 3 | Hooks | `~\u002F.claude\u002Fhooks\u002F` + settings | Scripts on disk referenced by no hook config; old versions superseded by rewrites |\n| 4 | Permissions | `permissions.allow` in `settings.json` \u002F `settings.local.json` | Duplicates; specific entries already covered by a wildcard; one-off grants from past experiments |\n| 5 | MCP servers | `~\u002F.claude.json`, project `.mcp.json` | Servers that fail to connect; functional duplicates; long unused |\n| 6 | Automations | scheduled jobs \u002F reminders, wherever kept | Fired one-shots older than 30 days; jobs whose target script no longer exists |\n| 7 | Project history | `~\u002F.claude\u002Fprojects\u002F*\u002F` | Stale handoff snapshots; session records superseded by newer state |\n| 8 | Runtime caches | `cache\u002F`, `file-history\u002F`, `logs\u002F`, `shell-snapshots\u002F` | Sort by size and mtime; propose items >30 days old and large |\n| 9 | Context cost | every loaded component | See §Context budget below — a component can be healthy and still not worth its tokens |\n\n## Config tamper guard (harvested from ECC `plankton-code-quality`)\n\nA distinct failure mode from all nine channels above: an agent facing a\nfailing lint\u002Fformat\u002Fbuild gate **loosens the config instead of fixing the\ncode** — relaxing a `.eslintrc` rule, widening `ruff.toml`'s ignore list,\ndisabling a `.shellcheckrc`\u002F`.yamllint`\u002F`.hadolint.yaml` check — so the gate\ngoes green without the underlying issue being fixed. This is metric-gaming,\nnot cleanup, and it is the same shape of problem the false-positive\u002Fgaming\npatterns in `gan-harness-edho-ferdian\u002Freferences\u002Floop-design-review.md`\nalready name for evaluator loops (\"Goodhart-gaming the verifier\" — the score\nclimbs while the artifact gets worse). Treat a lint\u002Fformat config edit found\nduring a scan as a high-confidence flag, not a routine change: check whether\nit was made *because* a check was failing at the time, and whether it\nnarrows coverage rather than reflecting a genuinely agreed convention (a\nnarrowing with an inline comment explaining the deliberate exception is a\ndocumented convention, not tampering — same test `language-code-review-\nedho-ferdian`'s Reflection gate already applies to a suppressed lens\nfinding). This is a detection responsibility for this skill's channel 3\u002F4\nscans, not a new channel — a tampered config is the same \"stale\u002Funreachable\ncorrectness\" concern as any other hook\u002Fpermission drift, just with a\ngaming motive behind it.\n\n## Language\u002Ftool gate table (harvested from ECC `plankton-code-quality`)\n\nUseful as a quick cross-check when this skill (or `build-fix-edho-ferdian`\nPhase 0) needs to confirm which formatter\u002Flinter *should* be running for a\ngiven file extension before treating its absence as a gap:\n\n| Extension \u002F stack | Formatter | Linter | Optional |\n|---|---|---|---|\n| `.py` | ruff format | ruff | ty, vulture, bandit |\n| `.ts`\u002F`.tsx`\u002F`.js`\u002F`.jsx` | biome (or prettier) | biome (or eslint) | oxlint, semgrep, knip |\n| `.sh`\u002F`.bash` | shfmt | shellcheck | — |\n| `.yml`\u002F`.yaml` | — | yamllint | — |\n| `.md` | — | markdownlint-cli2 | — |\n| `Dockerfile` | — | hadolint (≥2.12.0) | — |\n| `.toml` | taplo | — | — |\n| `.json` | jaq | — | — |\n\nNot exhaustive for every stack in this ecosystem (e.g. Go\u002FRust\u002FJava tooling\nlives in those stacks' own build-fix\u002Freview references) — this table covers\nwhat ECC's `plankton-code-quality` ships and is meant as a fast reference,\nnot a replacement for the per-stack detection tables already in\n`build-fix-edho-ferdian` and `language-code-review-edho-ferdian`. It is\nalso not the setup guide — installing and wiring any of these tools\n(including the Husky\u002Flint-staged git-hook layer around them) is\n`code-quality-tooling-edho-ferdian`'s job; this table exists only so this\nskill's own tamper-detection scan (above) has a fast fact-check for\n\"what should be running here\" before flagging its absence or a suspicious\nedit as drift.\n\n## Context budget (channel 9)\n\nChannels 1-8 ask \"is this dead?\". Channel 9 asks a different and often more\nuseful question: **\"is this alive but not worth what it costs to keep\nloaded?\"** A perfectly maintained skill that is loaded every session and\nfires twice a year is a context-budget problem, not a staleness problem.\n\nEstimate per component — skills (frontmatter always loaded, body on\ntrigger), MCP server tool definitions (always loaded, frequently the single\nlargest line item), rules and always-on memory files, hook definitions.\nRank by tokens-per-session against observed usefulness, and report the top\noffenders with a concrete recommendation: disable, narrow the description\nso it stops loading speculatively, or move to LIBRARY (below).\n\n**DAILY vs LIBRARY.** Borrowed from ECC `agent-sort` (fetched 2026-09-04) —\nthe classification idea only, never its ECC install-planning machinery.\nSplit components in two buckets, and require repo evidence for every DAILY\ncall rather than a feeling:\n\n- **DAILY** — should be available every session for the work actually being\n  done. Justified by concrete evidence: file extensions, lockfiles,\n  framework configs, CI config, imports.\n- **LIBRARY** — worth keeping reachable, not worth loading by default.\n  LIBRARY does not mean delete.\n\n## Strategic compaction\n\nNot everything here is deletion. Context also gets reclaimed by compacting\nat the right moment. Auto-compaction fires at an arbitrary token threshold,\nwhich usually lands mid-task and discards exactly the working state that\nwas still needed.\n\nCompact deliberately at a **phase boundary** instead — for work running\nunder `dev-kickoff-edho-ferdian`, that means after Stage 6 REMEMBER of a\ntask, once the memory files have been written and the state that matters\nis on disk rather than in the window. A session that has just snapshotted\ncan afford to lose its context; one mid-IMPLEMENT cannot.\n\n## Reference: harness operation\n\nHook wiring per-type and per-language, model routing tiers, context-budget\nmechanics, permission-widening posture, and todo-list-as-steering-surface —\nall detail that channels 3 (hooks), 6 (automations), and 9 (context cost)\ndraw on but that doesn't need to sit inline in this file — lives in\n`references\u002Fharness-operation.md`.\n\n## ECC decommissioning track\n\nThis ecosystem's stated end state (D-005) is to stop depending on the ECC\ninstall entirely. That removal is a config-hygiene operation, and it is the\none place where this skill is allowed to look at ECC-managed paths — to\nremove them, never to maintain them.\n\nRun this track only on explicit request. It is ordered so that nothing\nbreaks before its replacement exists:\n\n1. **Inventory what is actually ECC-managed** — `~\u002F.claude\u002Fagents\u002F*`,\n   `~\u002F.claude\u002Frules\u002Fecc\u002F*`, ECC-installed skills, ECC hook entries in\n   `settings.json`, ECC-specific env vars (`ECC_*`).\n2. **Map each one to its native replacement** in this ecosystem's own\n   `skills\u002F` (the porting record in `project-memory\u002F02-gap-analysis.md`\n   and the decision register are the source of truth for this map).\n3. **Report the unreplaced remainder.** Anything ECC-managed with no native\n   equivalent is a porting gap, not a deletion candidate — surface it and\n   stop. Do not remove a capability that has no replacement.\n4. **Disable before deleting.** Rename to `.disabled`, run normally for at\n   least one working session, and only then move to `_gc_trash\u002F`.\n5. **Check for dangling references** after each removal: grep this\n   ecosystem's own `skills\u002F` for pointers into ECC paths. A native skill\n   pointing at an ECC file is a D-005 violation and was already found once\n   during the kelompok 3 port (`e2e-testing-edho-ferdian`, 4 locations).\n\n## Workflow\n\n1. **Scan** the channels the user named, or all of them. Collect: path,\n   channel, the signal that flagged it, size, last-modified.\n2. **Rank by confidence** — broken\u002Forphaned is high, merely old is low —\n   and present a numbered table. Cap at ~20.\n3. **Confirm one by one.** Show the evidence, then `[y\u002Fn\u002Fskip]`. The user\n   can stop at any point.\n4. **Soft-delete** what was confirmed. Permission entries live in JSON with\n   no comments: back up the settings file, record each removed entry\n   verbatim in `gc_log.md`, then remove it from the `allow` array.\n5. **Log** the run.\n6. **Report**: what was reclaimed, which channels are healthy, suggested\n   next review date.\n\n## Rules\n\n- Never delete autonomously, and never offer a \"yes to all\".\n- An item that is merely old is a low-confidence candidate. Broken,\n  orphaned, or unreachable is high-confidence. Do not present them at the\n  same severity.\n- Do not touch project source code. That is refactoring — see\n  `dead-code-cleanup-edho-ferdian`.\n- Do not touch this ecosystem's `skills\u002F` content quality; that is\n  `skill-audit-edho-ferdian`'s job. Channel 1 here only asks whether a\n  skill is *installed and loaded*, never whether it is *well written*.\n- ECC-managed paths are in scope only inside the decommissioning track,\n  and only ever to remove them.\n\n## Provenance\n\nAdapted from ECC `config-gc`, `context-budget`, `workspace-surface-audit`,\n`automation-audit-ops`, `strategic-compact`, and the DAILY\u002FLIBRARY concept\nof `agent-sort` — all fetched 2026-09-04. Consolidated into one skill with\nper-channel lenses per D-009 rather than six near-overlapping skills.\nThe ECC decommissioning track has no upstream equivalent; it exists\nbecause of D-005 and inverts `agent-sort`'s purpose (which was to plan an\nECC install, not to retire one). The config tamper guard and language\u002Ftool\ngate table are a small harvest from ECC `plankton-code-quality`, fetched\n2026-09-06 — only those two concepts were pulled in, not its PreToolUse\u002F\nStop-hook enforcement machinery, since this skill works by human-confirmed\nscan-and-flag, not automatic blocking. `references\u002Fharness-operation.md`\nis a further D-005 harvest, fetched 2026-09-06, from ECC\n`rules\u002Fcommon\u002Fhooks.md`, `rules\u002Fcommon\u002Fperformance.md`, and the 21\n`rules\u002F\u003Clang>\u002Fhooks.md` files — these were still active global instructions\nloading from the ECC rule tree, and this skill already owns the `~\u002F.claude`\nsurface those files govern.\n","# Config Hygiene — Edho Ferdian Mode\n\n## Scope — read this first\n\n**Target: the user's own Claude Code environment** — `~\u002F.claude\u002F*`,\nproject-level `.claude\u002F*`, `~\u002F.claude.json`, and `.mcp.json`.\n\nThis is the deliberate complement to `skill-audit-edho-ferdian`, which\naudits the *content quality* of this ecosystem's `skills\u002F*` and explicitly\nrefuses to look at `~\u002F.claude`. This skill looks at exactly what that one\nrefuses to: the environment those skills run inside.\n\n| Question | Skill |\n|---|---|\n| Is this ecosystem's skill well-written, non-redundant, links intact? | `skill-audit-edho-ferdian` |\n| Should a new skill exist at all, and is it actually obeyed? | `skill-authoring-edho-ferdian` |\n| Is the environment carrying dead weight, and what does it cost? | **this skill** |\n\nNever delete autonomously. Every candidate gets its own confirmation.\n\n## Design principles\n\n1. **Append-only configs leak.** Skills, memory files, hooks and permission\n   entries only ever get added. Without a periodic pass they rot silently.\n2. **Periodic beats one-time.** Scan roughly every 30 days, propose a small\n   batch each time. Cap a run at ~20 candidates — this is GC, not a purge.\n3. **Per-channel signals.** Each accumulation type has its own staleness\n   test. Do not apply one rule everywhere.\n4. **Soft-delete first.** Rename to `.disabled` > move to\n   `~\u002F.claude\u002F_gc_trash\u002F\u003Cdate>\u002F` > real deletion. Always keep an undo path.\n5. **Human-in-the-loop, per item.** Show the evidence, then `[y\u002Fn\u002Fskip]`.\n   There is no \"yes to all\".\n6. **Log every run** to `~\u002F.claude\u002Fgc_log.md`: what was touched, why, and\n   how to undo it.\n\n## Scan channels\n\n| # | Channel | Path | Staleness \u002F redundancy signals |\n|---|---|---|---|\n| 1 | Skills | `~\u002F.claude\u002Fskills\u002F*\u002F` | Heavily overlapping names; never triggered in recent transcripts; domain mismatch with actual work; broken or empty SKILL.md |\n| 2 | Memory | `~\u002F.claude\u002F**\u002Fmemory\u002F*.md` + index | Several index entries for one topic; contents contradicted by newer entries; dates already passed; orphans missing from the index; sub-100-word fragments that should merge |\n| 3 | Hooks | `~\u002F.claude\u002Fhooks\u002F` + settings | Scripts on disk referenced by no hook config; old versions superseded by rewrites |\n| 4 | Permissions | `permissions.allow` in `settings.json` \u002F `settings.local.json` | Duplicates; specific entries already covered by a wildcard; one-off grants from past experiments |\n| 5 | MCP servers | `~\u002F.claude.json`, project `.mcp.json` | Servers that fail to connect; functional duplicates; long unused |\n| 6 | Automations | scheduled jobs \u002F reminders, wherever kept | Fired one-shots older than 30 days; jobs whose target script no longer exists |\n| 7 | Project history | `~\u002F.claude\u002Fprojects\u002F*\u002F` | Stale handoff snapshots; session records superseded by newer state |\n| 8 | Runtime caches | `cache\u002F`, `file-history\u002F`, `logs\u002F`, `shell-snapshots\u002F` | Sort by size and mtime; propose items >30 days old and large |\n| 9 | Context cost | every loaded component | See §Context budget below — a component can be healthy and still not worth its tokens |\n\n## Config tamper guard (harvested from ECC `plankton-code-quality`)\n\nA distinct failure mode from all nine channels above: an agent facing a\nfailing lint\u002Fformat\u002Fbuild gate **loosens the config instead of fixing the\ncode** — relaxing a `.eslintrc` rule, widening `ruff.toml`'s ignore list,\ndisabling a `.shellcheckrc`\u002F`.yamllint`\u002F`.hadolint.yaml` check — so the gate\ngoes green without the underlying issue being fixed. This is metric-gaming,\nnot cleanup, and it is the same shape of problem the false-positive\u002Fgaming\npatterns in `gan-harness-edho-ferdian\u002Freferences\u002Floop-design-review.md`\nalready name for evaluator loops (\"Goodhart-gaming the verifier\" — the score\nclimbs while the artifact gets worse). Treat a lint\u002Fformat config edit found\nduring a scan as a high-confidence flag, not a routine change: check whether\nit was made *because* a check was failing at the time, and whether it\nnarrows coverage rather than reflecting a genuinely agreed convention (a\nnarrowing with an inline comment explaining the deliberate exception is a\ndocumented convention, not tampering — same test `language-code-review-\nedho-ferdian`'s Reflection gate already applies to a suppressed lens\nfinding). This is a detection responsibility for this skill's channel 3\u002F4\nscans, not a new channel — a tampered config is the same \"stale\u002Funreachable\ncorrectness\" concern as any other hook\u002Fpermission drift, just with a\ngaming motive behind it.\n\n## Language\u002Ftool gate table (harvested from ECC `plankton-code-quality`)\n\nUseful as a quick cross-check when this skill (or `build-fix-edho-ferdian`\nPhase 0) needs to confirm which formatter\u002Flinter *should* be running for a\ngiven file extension before treating its absence as a gap:\n\n| Extension \u002F stack | Formatter | Linter | Optional |\n|---|---|---|---|\n| `.py` | ruff format | ruff | ty, vulture, bandit |\n| `.ts`\u002F`.tsx`\u002F`.js`\u002F`.jsx` | biome (or prettier) | biome (or eslint) | oxlint, semgrep, knip |\n| `.sh`\u002F`.bash` | shfmt | shellcheck | — |\n| `.yml`\u002F`.yaml` | — | yamllint | — |\n| `.md` | — | markdownlint-cli2 | — |\n| `Dockerfile` | — | hadolint (≥2.12.0) | — |\n| `.toml` | taplo | — | — |\n| `.json` | jaq | — | — |\n\nNot exhaustive for every stack in this ecosystem (e.g. Go\u002FRust\u002FJava tooling\nlives in those stacks' own build-fix\u002Freview references) — this table covers\nwhat ECC's `plankton-code-quality` ships and is meant as a fast reference,\nnot a replacement for the per-stack detection tables already in\n`build-fix-edho-ferdian` and `language-code-review-edho-ferdian`. It is\nalso not the setup guide — installing and wiring any of these tools\n(including the Husky\u002Flint-staged git-hook layer around them) is\n`code-quality-tooling-edho-ferdian`'s job; this table exists only so this\nskill's own tamper-detection scan (above) has a fast fact-check for\n\"what should be running here\" before flagging its absence or a suspicious\nedit as drift.\n\n## Context budget (channel 9)\n\nChannels 1-8 ask \"is this dead?\". Channel 9 asks a different and often more\nuseful question: **\"is this alive but not worth what it costs to keep\nloaded?\"** A perfectly maintained skill that is loaded every session and\nfires twice a year is a context-budget problem, not a staleness problem.\n\nEstimate per component — skills (frontmatter always loaded, body on\ntrigger), MCP server tool definitions (always loaded, frequently the single\nlargest line item), rules and always-on memory files, hook definitions.\nRank by tokens-per-session against observed usefulness, and report the top\noffenders with a concrete recommendation: disable, narrow the description\nso it stops loading speculatively, or move to LIBRARY (below).\n\n**DAILY vs LIBRARY.** Borrowed from ECC `agent-sort` (fetched 2026-09-04) —\nthe classification idea only, never its ECC install-planning machinery.\nSplit components in two buckets, and require repo evidence for every DAILY\ncall rather than a feeling:\n\n- **DAILY** — should be available every session for the work actually being\n  done. Justified by concrete evidence: file extensions, lockfiles,\n  framework configs, CI config, imports.\n- **LIBRARY** — worth keeping reachable, not worth loading by default.\n  LIBRARY does not mean delete.\n\n## Strategic compaction\n\nNot everything here is deletion. Context also gets reclaimed by compacting\nat the right moment. Auto-compaction fires at an arbitrary token threshold,\nwhich usually lands mid-task and discards exactly the working state that\nwas still needed.\n\nCompact deliberately at a **phase boundary** instead — for work running\nunder `dev-kickoff-edho-ferdian`, that means after Stage 6 REMEMBER of a\ntask, once the memory files have been written and the state that matters\nis on disk rather than in the window. A session that has just snapshotted\ncan afford to lose its context; one mid-IMPLEMENT cannot.\n\n## Reference: harness operation\n\nHook wiring per-type and per-language, model routing tiers, context-budget\nmechanics, permission-widening posture, and todo-list-as-steering-surface —\nall detail that channels 3 (hooks), 6 (automations), and 9 (context cost)\ndraw on but that doesn't need to sit inline in this file — lives in\n`references\u002Fharness-operation.md`.\n\n## ECC decommissioning track\n\nThis ecosystem's stated end state (D-005) is to stop depending on the ECC\ninstall entirely. That removal is a config-hygiene operation, and it is the\none place where this skill is allowed to look at ECC-managed paths — to\nremove them, never to maintain them.\n\nRun this track only on explicit request. It is ordered so that nothing\nbreaks before its replacement exists:\n\n1. **Inventory what is actually ECC-managed** — `~\u002F.claude\u002Fagents\u002F*`,\n   `~\u002F.claude\u002Frules\u002Fecc\u002F*`, ECC-installed skills, ECC hook entries in\n   `settings.json`, ECC-specific env vars (`ECC_*`).\n2. **Map each one to its native replacement** in this ecosystem's own\n   `skills\u002F` (the porting record in `project-memory\u002F02-gap-analysis.md`\n   and the decision register are the source of truth for this map).\n3. **Report the unreplaced remainder.** Anything ECC-managed with no native\n   equivalent is a porting gap, not a deletion candidate — surface it and\n   stop. Do not remove a capability that has no replacement.\n4. **Disable before deleting.** Rename to `.disabled`, run normally for at\n   least one working session, and only then move to `_gc_trash\u002F`.\n5. **Check for dangling references** after each removal: grep this\n   ecosystem's own `skills\u002F` for pointers into ECC paths. A native skill\n   pointing at an ECC file is a D-005 violation and was already found once\n   during the kelompok 3 port (`e2e-testing-edho-ferdian`, 4 locations).\n\n## Workflow\n\n1. **Scan** the channels the user named, or all of them. Collect: path,\n   channel, the signal that flagged it, size, last-modified.\n2. **Rank by confidence** — broken\u002Forphaned is high, merely old is low —\n   and present a numbered table. Cap at ~20.\n3. **Confirm one by one.** Show the evidence, then `[y\u002Fn\u002Fskip]`. The user\n   can stop at any point.\n4. **Soft-delete** what was confirmed. Permission entries live in JSON with\n   no comments: back up the settings file, record each removed entry\n   verbatim in `gc_log.md`, then remove it from the `allow` array.\n5. **Log** the run.\n6. **Report**: what was reclaimed, which channels are healthy, suggested\n   next review date.\n\n## Rules\n\n- Never delete autonomously, and never offer a \"yes to all\".\n- An item that is merely old is a low-confidence candidate. Broken,\n  orphaned, or unreachable is high-confidence. Do not present them at the\n  same severity.\n- Do not touch project source code. That is refactoring — see\n  `dead-code-cleanup-edho-ferdian`.\n- Do not touch this ecosystem's `skills\u002F` content quality; that is\n  `skill-audit-edho-ferdian`'s job. Channel 1 here only asks whether a\n  skill is *installed and loaded*, never whether it is *well written*.\n- ECC-managed paths are in scope only inside the decommissioning track,\n  and only ever to remove them.\n\n## Provenance\n\nAdapted from ECC `config-gc`, `context-budget`, `workspace-surface-audit`,\n`automation-audit-ops`, `strategic-compact`, and the DAILY\u002FLIBRARY concept\nof `agent-sort` — all fetched 2026-09-04. Consolidated into one skill with\nper-channel lenses per D-009 rather than six near-overlapping skills.\nThe ECC decommissioning track has no upstream equivalent; it exists\nbecause of D-005 and inverts `agent-sort`'s purpose (which was to plan an\nECC install, not to retire one). The config tamper guard and language\u002Ftool\ngate table are a small harvest from ECC `plankton-code-quality`, fetched\n2026-09-06 — only those two concepts were pulled in, not its PreToolUse\u002F\nStop-hook enforcement machinery, since this skill works by human-confirmed\nscan-and-flag, not automatic blocking. `references\u002Fharness-operation.md`\nis a further D-005 harvest, fetched 2026-09-06, from ECC\n`rules\u002Fcommon\u002Fhooks.md`, `rules\u002Fcommon\u002Fperformance.md`, and the 21\n`rules\u002F\u003Clang>\u002Fhooks.md` files — these were still active global instructions\nloading from the ECC rule tree, and this skill already owns the `~\u002F.claude`\nsurface those files govern.\n",{"name":5,"metadata":24,"description":7},{"last_reviewed":25},"2026-10-27","main","533d32ad55c92d40fbabe93cbe21fe048598fe5a","skills\u002Fconfig-hygiene-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 config-hygiene-edho-ferdian",{"tool":40,"command":41},"bunx","bunx skills add edhoferdian\u002FEEF --skill config-hygiene-edho-ferdian",{"tool":43,"command":44},"pnpm","pnpm dlx skills add edhoferdian\u002FEEF --skill config-hygiene-edho-ferdian","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fconfig-hygiene-edho-ferdian\u002FSKILL.md",true,[48,54,58,64],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"hooks\u002Flog-subagent-run.js","SCRIPT","脚本",3168,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fconfig-hygiene-edho-ferdian\u002Fhooks\u002Flog-subagent-run.js",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"hooks\u002Fstatusline-usage.js",3538,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fconfig-hygiene-edho-ferdian\u002Fhooks\u002Fstatusline-usage.js",{"path":59,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fharness-operation.md","REFERENCE","参考文档",20729,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fconfig-hygiene-edho-ferdian\u002Freferences\u002Fharness-operation.md",{"path":65,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":66,"sourceHtmlUrl":67},"scripts\u002Fsubagent-report.js",6550,"https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fconfig-hygiene-edho-ferdian\u002Fscripts\u002Fsubagent-report.js",[69,76,83,90,97,104],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"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":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"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":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"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":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"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":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M7C9F93YCXFCVRDG4S5QFAH4","skill-authoring-edho-ferdian","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\".","https:\u002F\u002Fgithub.com\u002Fedhoferdian\u002FEEF\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring-edho-ferdian\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"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",[],[]]