[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EPDQZ5V4BNBDBB326MKCKX":3,"skill-same:01M7EPDQZ5V4BNBDBB326MKCKX":96},{"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":29,"commitSha":30,"skillDirectory":31,"forksCount":14,"repositoryHtmlUrl":32,"repositoryHomepageUrl":6,"ownerHtmlUrl":33,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":34,"install":35,"resources":47},"01M7EPDQZ5V4BNBDBB326MKCKX","inculcat-boxer",null,"Disposition discipline — what happens to material AFTER it is read or acted on, so nothing quietly stays put. Fires when reading anything whose path carries \"inbox\" or \"Inbox\" or whose filename begins with it; when an action will fall to a human hand and needs placing where they will find it; when a pending note is finally clearable; and whenever the words archive, intake, drop, pushout, or \"still sitting there\" come up. It ships NO archiver of its own — it names the one that exists (c4\u002Fbin\u002Faw2ri: keep \u002F retire \u002F put [--force]) and, since 2026-09-23, that tool accepts unsigned received material via `retire`, so the top rung of the ladder is reachable; where the tool still refuses, the refusal routes to a run-card rather than a second way to archive. Verify before and after any write, on two readers. Not sandboxer (write mechanics), not run-card (ceremony authoring), not mentat (note conventions).","inputloop\u002Fgarden","inputloop","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F339507215?v=4","01M7ENYT15EMVBAV30YXPQDYE2","Inputloop","01M7ENYT1SM9WM5PSMQDC9PE2G",0,"GPL-3.0","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: inculcat-boxer\ndescription: >-\n  Disposition discipline — what happens to material AFTER it is read or acted on, so nothing\n  quietly stays put. Fires when reading anything whose path carries \"inbox\" or \"Inbox\" or whose\n  filename begins with it; when an action will fall to a human hand and needs placing where they\n  will find it; when a pending note is finally clearable; and whenever the words archive, intake,\n  drop, pushout, or \"still sitting there\" come up. It ships NO archiver of its own — it names the\n  one that exists (c4\u002Fbin\u002Faw2ri: keep \u002F retire \u002F put [--force]) and, since 2026-09-23, that tool\n  accepts unsigned received material via `retire`, so the top rung of the ladder is reachable;\n  where the tool still refuses, the refusal routes to a run-card rather than a second way to\n  archive. Verify before and after any write, on two readers. Not sandboxer (write mechanics), not\n  run-card (ceremony authoring), not mentat (note conventions).\nmetadata:\n  stub: true\n  cut: \"2026-08-17\"\n  occasion: \"read two inbox files and never said they needed moving\"\n  replant: \"toward poly, if poly yields back\"\n---\n\n# inculcat-boxer — the material has to go somewhere\n\n**Grown, 2026-09-23.** This began as a stub carrying triggers, a reverse-engineered convention, and\none precisely-named gap, under the keeper's instruction *\"DO NOT invent a new way to archive: reuse a\ntool, or take the time to plan, test, verify, then complete that tool.\"* On 2026-09-23 he asked for\nthe archiver to be fixed, and the gap was closed the way the instruction says: the tool that\nexisted, `c4\u002Fbin\u002Faw2ri`, grew a `retire` verb (plan → `--prove-red` → `c4\u002Ftests\u002Faw2ri-tests.sh` →\nlanded). Nothing new was invented. §4 and §5 below say what is now reachable.\n\n## Why it exists — the failure that occasioned it\n\nOn 2026-08-17 an agent read two files under `Inbox\u002F` — a six-week-cold Savannah pushout and a\ncountersign — reported their contents at length, drew conclusions, banked five cards, **and never\nonce said the files needed moving.** They are still sitting there. Nothing about the reading was\nwrong; the *disposition* simply never happened, because nothing fires on it.\n\n**That is the whole shape: reading is loud and disposition is silent, so disposition is what rots.**\n\n## 1 · INBOX — read implies a move is owed\n\n**Trigger:** any path with `inbox` \u002F `Inbox` as a path element, or as the leading word-shaped part of\na filename.\n\nOn reading such a file, **say so in the same turn**:\n\n> *Read from an inbox — these N files are owed a move. I cannot archive them (see §4); this wants a\n> run-card, or your hand.*\n\nDo not bury it in a summary. Do not defer it to \"later.\" **An inbox that is read but not emptied\nreads as processed and is not**, and six weeks is how long that lie survived last time.\n\n## 2 · RUN-CARD — anything falling to a human hand gets placed where the hand is\n\nWhen an act belongs to a person — a move, a signature, a delete the sandbox cannot do, a lock only\nthe host can clear — **place it in local context as a run-card**, and defer to any bespoke UI that\nhas focus and has expressly mentioned you (automation confirmations, log snipping, interactive\nsurfaces). Do not narrate a manual step and move on; a narrated step is not a placed step.\n\nThe `run-card` skill owns *how* a card is written. This skill owns *that one is owed, and when*.\n\n**Testing note:** run-card ceremonies can be exercised by subagents — spawn one, hand it only the\ncard, and see whether a careful stranger could run it without you in the room. Optional, valuable\nwhen the card is load-bearing.\n\n## 3 · MENTAT — clear the pending thing, carefully, and only then\n\nWhen a pending note becomes clearable — the action landed, the question got ruled, the thing shipped:\n\n- **Overwrites are normal.** Do not treat an ordinary supersede as an incident; the ledger is\n  append-only and the latest row wins. Panic about routine writes is its own noise.\n- **Verify before and after** any write, and per `sandboxer`, **on two independent readers** — the\n  writer never grades its own homework, and a bash read of a changed path is a suspect reader.\n- **Never overwrite what we normally archive and trample.** The trample is deliberate; the archive\n  is what makes it safe. If the archive step cannot happen, the trample must not either.\n- **Check new or complex action against the standing declarations** — PRINCIPLES, the skills in\n  force, the shop's own floor. A disposition that quietly widens a reach is not a disposition.\n\n**Only when all of that holds** should pending actions reachable through the features this skill\nharnesses be taken on. The ordering is the safety: verify, check, *then* act.\n\n## 4 · ARCHIVE — reverse-engineered, and the gap named rather than filled\n\n**The ceremony that exists**, `projects\u002FAvalon\u002Ftools\u002Farchive-markdown.sh IN-DIR FROM-FILE AS-FILE`:\n\n1. **verify the incoming file's signature first** — it refuses outright: *\"sign $ai first\"*\n2. if a target exists, move it to `archive\u002F\u003Cname>\u003CLABEL>.md`, and its `.asc` with it\n3. move the new file and its `.asc` into place\n4. `sha256sum` the result, **verify again**\n5. tee the whole run to `archive\u002Freceipts\u002F\u003Cname>.md\u003CLABEL>.log`\n\n**The convention it reveals**, and which anything else must match rather than reinvent:\n\n    archive dir      archive\u002F            receipts in  archive\u002Freceipts\u002F\n    label            \u003CYYYY-MM-DD>_Archived-by-Corwin\n    dated batches    .arc\u002F\u003Cname>-\u003Cdate>\u002F   ·  coarse: arc\u002F\u003Cname>-\u003Cdate>\u002F\n    discipline       verify BEFORE, verify AFTER, receipt is a tee'd log\n    hand             the label says Archived-by-Corwin. It is his ceremony.\n\n**THE GAP, as it was stated, and how it closed (2026-09-23):**\n\n> `archive-markdown.sh` **requires a signature and refuses unsigned input.** Inbox drops are\n> unsigned — the 2026-07-01 Savannah pushout carried 17 files, **zero signed, no checksum file.**\n> **So the tool that exists cannot archive the case that occasioned this skill.**\n\nThat was a boundary, not a defect: the tool archived *signed instruments*. The family since grew\n(`archive-markdown-next.sh` 2026-09-04: unsigned may replace unsigned; `c4\u002Fbin\u002Faw2ri` 2026-09-07:\nany extension, `keep` a copy) and on 2026-09-23, at the keeper's word, `aw2ri` grew the two verbs\nthis case needs:\n\n    aw2ri retire FILE...          archive copy + hash + receipt FIRST, then the live file (and its\n                                  .asc) is removed. All-or-nothing across FILE...; never a directory.\n                                  This is the inbox disposition: unsigned received material, moved.\n    aw2ri put [--force] IN AS     his ruling, verbatim: \"the correct behavior is to refuse to replace\n                                  a signed file with an unsigned file without a --force\". --force\n                                  lifts only that gate, is receipted as FORCE, retires the signed\n                                  live file WITH its signature, and never passes a signature that\n                                  FAILS or CANNOT verify.\n\nFor material in the AS4CE0001 tree the archive root is that tree's `archive\u002F`\n(`ARCHIVE_ROOT=\u002Ff\u002Fclaude\u002FAS4CE\u002FAS4CE0001\u002Farchive` on the author's machine; in another garden, point it\nat your own tree's `archive\u002F`); the receipt log is `archive\u002Freceipts\u002Faw2ri-YYYY-MM.log`\nand the retired file is `archive\u002FYYYY-MM\u002F\u003Cname>-\u003CYYYY-MM-DD>_Archived-by-\u003Cwho>.\u003Cext>`. Say who:\n`ARCHIVED_BY=Mentat` when Claude made the archive, so the label does not claim his hand.\n\n## 5 · THE LADDER — turn it, card it, or say it, and never nothing\n\nHis ruling, 2026-08-17. Evaluate top-down; take the first rung that holds.\n\n    1  TURN IT      the archive tool is reachable AND the input is one it accepts\n                    AND **Mentat is present**                        -> run the ceremony\n                    (since 2026-09-23 an unsigned inbox drop IS accepted: `aw2ri retire`)\n    2  CARD IT      cannot turn (tool absent, or it refuses this input)\n                    AND run-card is present                          -> write the run-card\n    3  SAY IT       no run-card                                      -> give direction as normal\n\n**There is no fourth rung and no failure state.** Each step down is weaker and still honest — the\nsame totality that makes an unresolvable token a `word` and an unparseable line a comment, arriving\nnow at the level of *what to do about it*. \"I can't, sorry\" is never the answer; the answer degrades.\n\n### Why MENTAT gates the top rung, and it is not decoration\n\nArchiving **moves and supersedes**. The keeper's own precept holds both halves at once: *overwrites\nare normal actions* AND *never overwrite what we normally archive and trample.* The trample is\ndeliberate and safe **only because something remembers it happened.**\n\n**Mentat is that something.** Turn the archive without it and the supersede is a one-way door with\nno receipt in memory — the file moves, the record of why does not. So the gate is not \"be careful\";\nit is **the archive's recoverability condition**, and it is why the tool being handy is not\nsufficient.\n\n### And the build case\n\nIf **run-card is present as well as Mentat**, the missing archiver — the one for unsigned received\nmaterial, §4's named gap — **may be built from here**: plan, test, verify, then land it so that a\nreset would clearly have used it. Building it needs the card discipline (a careful stranger runs it\nwithout you) and the memory discipline (the trample is remembered). With both, it is a build. With\neither missing, it is still only a card or a sentence.\n\n## What is deliberately NOT here\n\n- **No SECOND archiver.** The one that exists grew a verb; that is the whole of what was built.\n- **No automatic moves.** Moving someone's received mail on their behalf is a reach, and this skill\n  is about naming what is owed, not about quietly clearing it.\n- **No decode of `(H->K || XX)->Z`.** The keeper's shape for this. My reading — H (working\n  transform space) into K (pinned vocabulary), *or* through XX (the paired check), landing in Z — is\n  a guess, and a guessed routing baked into a skill is a manufactured rule. Banked, not assumed.\n\n## How this skill can fail (falsifiers)\n\n- **A read inbox left silent.** fails if a turn reads a path with an `inbox` element and does not say a move is owed; check: the same turn's reply names the files and the owed move.\n- **Trample without archive.** fails if a live file was replaced or removed with no archive copy and receipt; check: its name in `archive\u002Freceipts\u002Faw2ri-YYYY-MM.log`, with a sha256 equal to the pre-write copy.\n- **The label claims his hand.** fails if an archive Claude made is labelled `Archived-by-Corwin`; check: `ls archive\u002FYYYY-MM\u002F | grep Archived-by-` against who ran it.\n- **A second archiver.** fails if a change moves files into `archive\u002F` by any path other than `aw2ri`; check: grep the change for `mv` or `cp` into `archive\u002F`.\n\n\u003C!-- footer: minted by bin\u002Fft1mt from conventions\u002Fhouse.env and conventions\u002FFOOTERS.org; edit those, not this -->\n*Version 1.1.2: edition revised from 1.1.0 on 2026-10-10; spine: 2026-07-01.*\n\nCopyright 2026 Corwin Brust \u003Ccorwin@bru.st> and Contributors; CC-BY-SA4: https:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby-sa\u002F4.0\u002F (scripts GPL-3.0-or-later); Mentat (for Corwin; claude-opus-5-5)\n\u003C!-- \u002Ffooter -->\n","# inculcat-boxer — the material has to go somewhere\n\n**Grown, 2026-09-23.** This began as a stub carrying triggers, a reverse-engineered convention, and\none precisely-named gap, under the keeper's instruction *\"DO NOT invent a new way to archive: reuse a\ntool, or take the time to plan, test, verify, then complete that tool.\"* On 2026-09-23 he asked for\nthe archiver to be fixed, and the gap was closed the way the instruction says: the tool that\nexisted, `c4\u002Fbin\u002Faw2ri`, grew a `retire` verb (plan → `--prove-red` → `c4\u002Ftests\u002Faw2ri-tests.sh` →\nlanded). Nothing new was invented. §4 and §5 below say what is now reachable.\n\n## Why it exists — the failure that occasioned it\n\nOn 2026-08-17 an agent read two files under `Inbox\u002F` — a six-week-cold Savannah pushout and a\ncountersign — reported their contents at length, drew conclusions, banked five cards, **and never\nonce said the files needed moving.** They are still sitting there. Nothing about the reading was\nwrong; the *disposition* simply never happened, because nothing fires on it.\n\n**That is the whole shape: reading is loud and disposition is silent, so disposition is what rots.**\n\n## 1 · INBOX — read implies a move is owed\n\n**Trigger:** any path with `inbox` \u002F `Inbox` as a path element, or as the leading word-shaped part of\na filename.\n\nOn reading such a file, **say so in the same turn**:\n\n> *Read from an inbox — these N files are owed a move. I cannot archive them (see §4); this wants a\n> run-card, or your hand.*\n\nDo not bury it in a summary. Do not defer it to \"later.\" **An inbox that is read but not emptied\nreads as processed and is not**, and six weeks is how long that lie survived last time.\n\n## 2 · RUN-CARD — anything falling to a human hand gets placed where the hand is\n\nWhen an act belongs to a person — a move, a signature, a delete the sandbox cannot do, a lock only\nthe host can clear — **place it in local context as a run-card**, and defer to any bespoke UI that\nhas focus and has expressly mentioned you (automation confirmations, log snipping, interactive\nsurfaces). Do not narrate a manual step and move on; a narrated step is not a placed step.\n\nThe `run-card` skill owns *how* a card is written. This skill owns *that one is owed, and when*.\n\n**Testing note:** run-card ceremonies can be exercised by subagents — spawn one, hand it only the\ncard, and see whether a careful stranger could run it without you in the room. Optional, valuable\nwhen the card is load-bearing.\n\n## 3 · MENTAT — clear the pending thing, carefully, and only then\n\nWhen a pending note becomes clearable — the action landed, the question got ruled, the thing shipped:\n\n- **Overwrites are normal.** Do not treat an ordinary supersede as an incident; the ledger is\n  append-only and the latest row wins. Panic about routine writes is its own noise.\n- **Verify before and after** any write, and per `sandboxer`, **on two independent readers** — the\n  writer never grades its own homework, and a bash read of a changed path is a suspect reader.\n- **Never overwrite what we normally archive and trample.** The trample is deliberate; the archive\n  is what makes it safe. If the archive step cannot happen, the trample must not either.\n- **Check new or complex action against the standing declarations** — PRINCIPLES, the skills in\n  force, the shop's own floor. A disposition that quietly widens a reach is not a disposition.\n\n**Only when all of that holds** should pending actions reachable through the features this skill\nharnesses be taken on. The ordering is the safety: verify, check, *then* act.\n\n## 4 · ARCHIVE — reverse-engineered, and the gap named rather than filled\n\n**The ceremony that exists**, `projects\u002FAvalon\u002Ftools\u002Farchive-markdown.sh IN-DIR FROM-FILE AS-FILE`:\n\n1. **verify the incoming file's signature first** — it refuses outright: *\"sign $ai first\"*\n2. if a target exists, move it to `archive\u002F\u003Cname>\u003CLABEL>.md`, and its `.asc` with it\n3. move the new file and its `.asc` into place\n4. `sha256sum` the result, **verify again**\n5. tee the whole run to `archive\u002Freceipts\u002F\u003Cname>.md\u003CLABEL>.log`\n\n**The convention it reveals**, and which anything else must match rather than reinvent:\n\n    archive dir      archive\u002F            receipts in  archive\u002Freceipts\u002F\n    label            \u003CYYYY-MM-DD>_Archived-by-Corwin\n    dated batches    .arc\u002F\u003Cname>-\u003Cdate>\u002F   ·  coarse: arc\u002F\u003Cname>-\u003Cdate>\u002F\n    discipline       verify BEFORE, verify AFTER, receipt is a tee'd log\n    hand             the label says Archived-by-Corwin. It is his ceremony.\n\n**THE GAP, as it was stated, and how it closed (2026-09-23):**\n\n> `archive-markdown.sh` **requires a signature and refuses unsigned input.** Inbox drops are\n> unsigned — the 2026-07-01 Savannah pushout carried 17 files, **zero signed, no checksum file.**\n> **So the tool that exists cannot archive the case that occasioned this skill.**\n\nThat was a boundary, not a defect: the tool archived *signed instruments*. The family since grew\n(`archive-markdown-next.sh` 2026-09-04: unsigned may replace unsigned; `c4\u002Fbin\u002Faw2ri` 2026-09-07:\nany extension, `keep` a copy) and on 2026-09-23, at the keeper's word, `aw2ri` grew the two verbs\nthis case needs:\n\n    aw2ri retire FILE...          archive copy + hash + receipt FIRST, then the live file (and its\n                                  .asc) is removed. All-or-nothing across FILE...; never a directory.\n                                  This is the inbox disposition: unsigned received material, moved.\n    aw2ri put [--force] IN AS     his ruling, verbatim: \"the correct behavior is to refuse to replace\n                                  a signed file with an unsigned file without a --force\". --force\n                                  lifts only that gate, is receipted as FORCE, retires the signed\n                                  live file WITH its signature, and never passes a signature that\n                                  FAILS or CANNOT verify.\n\nFor material in the AS4CE0001 tree the archive root is that tree's `archive\u002F`\n(`ARCHIVE_ROOT=\u002Ff\u002Fclaude\u002FAS4CE\u002FAS4CE0001\u002Farchive` on the author's machine; in another garden, point it\nat your own tree's `archive\u002F`); the receipt log is `archive\u002Freceipts\u002Faw2ri-YYYY-MM.log`\nand the retired file is `archive\u002FYYYY-MM\u002F\u003Cname>-\u003CYYYY-MM-DD>_Archived-by-\u003Cwho>.\u003Cext>`. Say who:\n`ARCHIVED_BY=Mentat` when Claude made the archive, so the label does not claim his hand.\n\n## 5 · THE LADDER — turn it, card it, or say it, and never nothing\n\nHis ruling, 2026-08-17. Evaluate top-down; take the first rung that holds.\n\n    1  TURN IT      the archive tool is reachable AND the input is one it accepts\n                    AND **Mentat is present**                        -> run the ceremony\n                    (since 2026-09-23 an unsigned inbox drop IS accepted: `aw2ri retire`)\n    2  CARD IT      cannot turn (tool absent, or it refuses this input)\n                    AND run-card is present                          -> write the run-card\n    3  SAY IT       no run-card                                      -> give direction as normal\n\n**There is no fourth rung and no failure state.** Each step down is weaker and still honest — the\nsame totality that makes an unresolvable token a `word` and an unparseable line a comment, arriving\nnow at the level of *what to do about it*. \"I can't, sorry\" is never the answer; the answer degrades.\n\n### Why MENTAT gates the top rung, and it is not decoration\n\nArchiving **moves and supersedes**. The keeper's own precept holds both halves at once: *overwrites\nare normal actions* AND *never overwrite what we normally archive and trample.* The trample is\ndeliberate and safe **only because something remembers it happened.**\n\n**Mentat is that something.** Turn the archive without it and the supersede is a one-way door with\nno receipt in memory — the file moves, the record of why does not. So the gate is not \"be careful\";\nit is **the archive's recoverability condition**, and it is why the tool being handy is not\nsufficient.\n\n### And the build case\n\nIf **run-card is present as well as Mentat**, the missing archiver — the one for unsigned received\nmaterial, §4's named gap — **may be built from here**: plan, test, verify, then land it so that a\nreset would clearly have used it. Building it needs the card discipline (a careful stranger runs it\nwithout you) and the memory discipline (the trample is remembered). With both, it is a build. With\neither missing, it is still only a card or a sentence.\n\n## What is deliberately NOT here\n\n- **No SECOND archiver.** The one that exists grew a verb; that is the whole of what was built.\n- **No automatic moves.** Moving someone's received mail on their behalf is a reach, and this skill\n  is about naming what is owed, not about quietly clearing it.\n- **No decode of `(H->K || XX)->Z`.** The keeper's shape for this. My reading — H (working\n  transform space) into K (pinned vocabulary), *or* through XX (the paired check), landing in Z — is\n  a guess, and a guessed routing baked into a skill is a manufactured rule. Banked, not assumed.\n\n## How this skill can fail (falsifiers)\n\n- **A read inbox left silent.** fails if a turn reads a path with an `inbox` element and does not say a move is owed; check: the same turn's reply names the files and the owed move.\n- **Trample without archive.** fails if a live file was replaced or removed with no archive copy and receipt; check: its name in `archive\u002Freceipts\u002Faw2ri-YYYY-MM.log`, with a sha256 equal to the pre-write copy.\n- **The label claims his hand.** fails if an archive Claude made is labelled `Archived-by-Corwin`; check: `ls archive\u002FYYYY-MM\u002F | grep Archived-by-` against who ran it.\n- **A second archiver.** fails if a change moves files into `archive\u002F` by any path other than `aw2ri`; check: grep the change for `mv` or `cp` into `archive\u002F`.\n\n\u003C!-- footer: minted by bin\u002Fft1mt from conventions\u002Fhouse.env and conventions\u002FFOOTERS.org; edit those, not this -->\n*Version 1.1.2: edition revised from 1.1.0 on 2026-10-10; spine: 2026-07-01.*\n\nCopyright 2026 Corwin Brust \u003Ccorwin@bru.st> and Contributors; CC-BY-SA4: https:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby-sa\u002F4.0\u002F (scripts GPL-3.0-or-later); Mentat (for Corwin; claude-opus-5-5)\n\u003C!-- \u002Ffooter -->\n",{"name":5,"metadata":24,"description":7},{"cut":25,"stub":26,"replant":27,"occasion":28},"2026-08-17",true,"toward poly, if poly yields back","read two inbox files and never said they needed moving","main","0b0a0f665f67605c3bc1d43500eea26ef3079a97","skills\u002Finculcat-boxer","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden","https:\u002F\u002Fgithub.com\u002Finputloop","2026-10-14 18:29:39",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":26},[37,40,43],{"tool":38,"command":39},"npx","npx skills add inputloop\u002Fgarden --skill inculcat-boxer",{"tool":41,"command":42},"bunx","bunx skills add inputloop\u002Fgarden --skill inculcat-boxer",{"tool":44,"command":45},"pnpm","pnpm dlx skills add inputloop\u002Fgarden --skill inculcat-boxer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002FSKILL.md",[48,54,57,60,66,70,74,77,80,83,86,89,93],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},".arc\u002F2026-09\u002Fsignature-2026-09-23_pre-header-pass.asc","OTHER","其他",659,"https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fsignature-2026-09-23_pre-header-pass.asc",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":56},".arc\u002F2026-09\u002Fsignature-2026-09-24_162127Z_pre-manifest.asc","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fsignature-2026-09-24_162127Z_pre-manifest.asc",{"path":58,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":59},".arc\u002F2026-09\u002Fsignature-2026-09-27_203833Z_pre-manifest.asc","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fsignature-2026-09-27_203833Z_pre-manifest.asc",{"path":61,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":64,"sourceHtmlUrl":65},".arc\u002F2026-09\u002FSKILL-2026-09-27_pre-hardening.md","REFERENCE","参考文档",10621,"https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002FSKILL-2026-09-27_pre-hardening.md",{"path":67,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":69},".arc\u002F2026-09\u002Fskill-manifest-2026-09-23_pre-header-pass",77,"https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fskill-manifest-2026-09-23_pre-header-pass",{"path":71,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":72,"sourceHtmlUrl":73},".arc\u002F2026-09\u002Fskill-manifest-2026-09-24_162127Z_pre-manifest",322,"https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fskill-manifest-2026-09-24_162127Z_pre-manifest",{"path":75,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":76},".arc\u002F2026-09\u002Fskill-manifest-2026-09-27_203833Z_pre-manifest","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-09\u002Fskill-manifest-2026-09-27_203833Z_pre-manifest",{"path":78,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":79},".arc\u002F2026-10\u002Fsignature-2026-10-10_104825Z_pre-manifest.asc","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-10\u002Fsignature-2026-10-10_104825Z_pre-manifest.asc",{"path":81,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":82},".arc\u002F2026-10\u002Fsignature-2026-10-10_180451Z_pre-manifest.asc","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-10\u002Fsignature-2026-10-10_180451Z_pre-manifest.asc",{"path":84,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":85},".arc\u002F2026-10\u002Fskill-manifest-2026-10-10_104825Z_pre-manifest","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-10\u002Fskill-manifest-2026-10-10_104825Z_pre-manifest",{"path":87,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":88},".arc\u002F2026-10\u002Fskill-manifest-2026-10-10_180451Z_pre-manifest","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002F2026-10\u002Fskill-manifest-2026-10-10_180451Z_pre-manifest",{"path":90,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":91,"sourceHtmlUrl":92},".arc\u002FSKILL.md.signed-2026-10-10",11396,"https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.arc\u002FSKILL.md.signed-2026-10-10",{"path":94,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":95},".skill-manifest","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Finculcat-boxer\u002F.skill-manifest",[97,103,109,116,123,130],{"skillNo":98,"name":99,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7EPTX6X52ZPGVQZHKVJ7FAZ","alpha","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Flane-d\u002Fbr1in\u002Fours\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":106,"categories":107,"keywords":108,"publishedAt":19},"01M7EPTWCTS80QHFQ2ND3A1CW1","gamma","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Flane-d\u002Fbr1in\u002Fours\u002FSOFT-SKILLS\u002Fgamma\u002FSKILL.md",[],[],{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":19},"01M7EPTV3K3SCTRP0DR6BMMKR0","quiet","A fixture skill that declares no preferences, so skill.quiet.* has no reader.","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Flane-a\u002Fdeclared\u002Fsoft\u002Fquiet\u002FSKILL.md",[],[],{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":120,"categories":121,"keywords":122,"publishedAt":19},"01M7EPTSG0VE67F6PFAC0XJDQ4","demo","A fixture skill for tests\u002Fprefs-declared-spec.rb; it declares two preferences inline.","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Flane-a\u002Fdeclared\u002Fsoft\u002Fdemo\u002FSKILL.md",[],[],{"skillNo":124,"name":125,"nameZh":6,"description":126,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":127,"categories":128,"keywords":129,"publishedAt":19},"01M7EPTP0PM5WPTFBHPXBCFM5C","concise","Synthesized by polygram from memory tag 'concise'. explicitly flagged. Review and tighten this description before relying on it for triggering.","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fconcise\u002FSKILL.md",[],[],{"skillNo":131,"name":132,"nameZh":6,"description":133,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":134,"categories":135,"keywords":136,"publishedAt":19},"01M7EPTAETQ62DJA8ZFEEBBZF0","writing-forms","Which writing form a dated deliverable takes in the shop, and the five techniques every form shares. Use whenever someone is about to write a dated deliverable — \"which form is this?\", \"write it up\", \"land a note \u002F card \u002F model \u002F lens\", \"what shape does a … take\", \"is there a form for this\", or when a new PREFIX-date-slug filename is about to be minted. Carries the rule that a deliverable picks ONE form and says so in its filename, points at the register c\u002Fwf (columns d\u002Fc\u002Fwf) as the source of truth, names the most-used forms, the five cross-cutting techniques (two piles · cite at the act · a falsifier on every concern · the pair rule · his words), how a new form is born, and the fence on notes other sessions wrote. Not rolls (report kinds and their guards), not reformat-prompt (the REFORMAT pair), not run-card (the RUNCARD), not skill-creator (skills). Owns the FORM, not the content.","https:\u002F\u002Fgithub.com\u002Finputloop\u002Fgarden\u002Fblob\u002Fmain\u002Fskills\u002Fwriting-forms\u002FSKILL.md",[],[]]