[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M75V1N9R2CHY0M8JMERZ59CS":3,"skill-same:01M75V1N9R2CHY0M8JMERZ59CS":46},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M75V1N9R2CHY0M8JMERZ59CS","glyphsteer-dense",null,"Use when running GlyphSteer's DENSE regime — embedding-based emoji steering via the containerized ONNX sidecar. Covers standing up the glyphsteer-serve Docker image, the MANDATORY emoji-collapse gate (many models map all emoji to one token), the magnitude probe, and the anchor metric-reshape (steer.sim_eff). Triggers on \"glyphsteer dense\", \"embedding steering\", \"emoji direction\", \"glyphsteer sidecar\", \"magnitude probe\", \"anchor reshape\", \"is the dense nudge big enough\".","sancovp\u002Fchaincompiler","sancovp","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F147458096?v=4","01M74F2EFE4Q0KF1Y5AY27HVNB","Isaac Wostrel-Rubin","01M75V171K7FNNX91AY22XG37A",0,"MIT","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fglyphsteer-dense\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: glyphsteer-dense\ndescription: Use when running GlyphSteer's DENSE regime — embedding-based emoji steering via the containerized ONNX sidecar. Covers standing up the glyphsteer-serve Docker image, the MANDATORY emoji-collapse gate (many models map all emoji to one token), the magnitude probe, and the anchor metric-reshape (steer.sim_eff). Triggers on \"glyphsteer dense\", \"embedding steering\", \"emoji direction\", \"glyphsteer sidecar\", \"magnitude probe\", \"anchor reshape\", \"is the dense nudge big enough\".\n---\n\n# GlyphSteer — the dense regime (containerized)\n\nThe dense half steers retrieval by the emoji's **learned direction** in embedding space.\nThe heavy neural runtime lives in a Docker sidecar (`fastembed → onnxruntime`, **no torch**);\nthe host calls it over HTTP and the steer\u002Fprobe math runs host-side in pure numpy.\n\n**Run its Python** from a plugin install with `uv run --no-project --with glyphsteer==0.3.6 python3` — pipe a block in as `python3 - \u003C\u003C'PY'` … `PY` — and from a clone of the public repository where `.\u002Finstall.sh` ran, with the clone's `python3`. For the local fastembed backend instead of the sidecar, run it with `uv run --no-project --with \"glyphsteer[dense]==0.3.6\" python3`.\n\n> The sidecar's compose file and the probes below are in a clone of the public repository: run every\n> `packages\u002F...` command from the clone's root (`...\u002Fchaincompiler`).\n\n## 1. Stand up the sidecar\n\n```bash\ncd packages\u002Fglyphsteer\u002Fserve\n# byte-aware multilingual model — REQUIRED (see the collapse gate below)\nGLYPHSTEER_MODEL=sentence-transformers\u002Fparaphrase-multilingual-MiniLM-L12-v2 \\\n  docker compose up --build -d\ncurl localhost:8088\u002Fhealth      # {\"status\":\"ok\",\"model\":\"...MiniLM-L12-v2\"}\ncurl localhost:8088\u002Finfo        # {\"model\":\"...\",\"dim\":384}\nexport GLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088   # host now uses the sidecar\n```\n\nThe host `dense.embed(texts)` auto-routes to the container when `GLYPHSTEER_EMBED_URL` is set\n(stdlib urllib — no host deps). Backends: `remote` (container, default when the env var is\nset), `fastembed` (local ONNX), `st` (local sentence-transformers).\n\n## 2. THE GATE — check emoji-collapse FIRST (non-negotiable)\n\nMany English tokenizers (BERT\u002FWordPiece, e.g. `bge-small-en-v1.5`) map **every** emoji to the\nsame token (`[UNK]`) → all emoji embed identically → dense steering is **impossible**. Always\nrun the gate before trusting any dense number:\n\n```bash\nGLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088 \\\n  python3 packages\u002Fglyphsteer\u002Fexperiments\u002Femoji_collapse_check.py\n# mean off-diagonal cosine ≈ 1.000 ⇒ COLLAPSED (reject the model)\n# well below 1.0 (e.g. 0.77–0.97) ⇒ DISTINCT (proceed)\n```\n\nVerified: `bge-small-en-v1.5` → 1.000 (blind). `paraphrase-multilingual-MiniLM-L12-v2` →\ndistinct, with clean alignment. Fix a collapse with a one-env-var model swap + rebuild.\n\n## 3. Magnitude probe (Q1–Q3)\n\n```bash\nGLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088 \\\n  python3 packages\u002Fglyphsteer\u002Fexperiments\u002Fmagnitude_probe.py\n```\n- **Q1 (alignment):** `cos(glyph, aligned_word) − cos(glyph, unrelated)` — should be clearly\n  positive (the emoji means what you think). Measured +0.28…+0.58 with the multilingual model.\n- **Q2 (per-token nudge):** injecting one emoji shifts a neutral chunk's vector by Δcos\n  ~−0.03…+0.09 — real but modest, and a single token can wash out or slightly *invert* under\n  mean-pooling (e.g. ❓ measured −0.034). This is why Q2 is a tuning concern, not a guarantee.\n- **Q3 (anchor reshape):** a controllable gain (×2 at weight 1).\n\n## 4. Steer — the anchor metric reshape (pure numpy, host-side)\n\n`sim_eff` (lives in `glyphsteer.steer`, re-exported at top level — `from glyphsteer import\nsim_eff`) bends the similarity metric toward chosen anchors at query time:\n`sim_eff(q,d) = cos(q,d) + Σ_k w_k·⟨q,â_k⟩·⟨d,â_k⟩`, with â_k from `dense.measure_anchor`.\n\n```python\nfrom glyphsteer import dense, sim_eff\nanchor = dense.measure_anchor(\"😡\")     # the glyph's learned direction (via the sidecar)\nq = dense.embed([\"the deploy broke\"])[0]\ndocs = dense.embed([\"it crashed badly\", \"it shipped fine\"])\nscored = sim_eff(q, docs, anchors=anchor[None, :], weights=[1.5])   # weight = the steering knob\n```\nweight 0 = plain cosine; raising it makes that axis dominate (too high → ignores the query).\n\n## Honest status\nDense is **viable and measured** but its per-token strength is a *tuning knob*, not a free\nlunch; end-to-end dense retrieval lift on a real (non-toy) corpus is the open item. Lexical\nfaceting (the core **glyphsteer** skill) is the regime that's a clean win today.\n\n## Verify (run this)\n```bash\ncd packages\u002Fglyphsteer\u002Fserve && docker compose up --build -d && cd -\nexport GLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088\npython3 packages\u002Fglyphsteer\u002Fexperiments\u002Femoji_collapse_check.py   # gate\npython3 packages\u002Fglyphsteer\u002Fexperiments\u002Fmagnitude_probe.py        # Q1–Q3\n```\nArchitecture\u002Fflows: `packages\u002Fglyphsteer\u002F.claude\u002Frules\u002F{10,20}-architecture-*.md` (DW-6\u002FDW-7).\n","# GlyphSteer — the dense regime (containerized)\n\nThe dense half steers retrieval by the emoji's **learned direction** in embedding space.\nThe heavy neural runtime lives in a Docker sidecar (`fastembed → onnxruntime`, **no torch**);\nthe host calls it over HTTP and the steer\u002Fprobe math runs host-side in pure numpy.\n\n**Run its Python** from a plugin install with `uv run --no-project --with glyphsteer==0.3.6 python3` — pipe a block in as `python3 - \u003C\u003C'PY'` … `PY` — and from a clone of the public repository where `.\u002Finstall.sh` ran, with the clone's `python3`. For the local fastembed backend instead of the sidecar, run it with `uv run --no-project --with \"glyphsteer[dense]==0.3.6\" python3`.\n\n> The sidecar's compose file and the probes below are in a clone of the public repository: run every\n> `packages\u002F...` command from the clone's root (`...\u002Fchaincompiler`).\n\n## 1. Stand up the sidecar\n\n```bash\ncd packages\u002Fglyphsteer\u002Fserve\n# byte-aware multilingual model — REQUIRED (see the collapse gate below)\nGLYPHSTEER_MODEL=sentence-transformers\u002Fparaphrase-multilingual-MiniLM-L12-v2 \\\n  docker compose up --build -d\ncurl localhost:8088\u002Fhealth      # {\"status\":\"ok\",\"model\":\"...MiniLM-L12-v2\"}\ncurl localhost:8088\u002Finfo        # {\"model\":\"...\",\"dim\":384}\nexport GLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088   # host now uses the sidecar\n```\n\nThe host `dense.embed(texts)` auto-routes to the container when `GLYPHSTEER_EMBED_URL` is set\n(stdlib urllib — no host deps). Backends: `remote` (container, default when the env var is\nset), `fastembed` (local ONNX), `st` (local sentence-transformers).\n\n## 2. THE GATE — check emoji-collapse FIRST (non-negotiable)\n\nMany English tokenizers (BERT\u002FWordPiece, e.g. `bge-small-en-v1.5`) map **every** emoji to the\nsame token (`[UNK]`) → all emoji embed identically → dense steering is **impossible**. Always\nrun the gate before trusting any dense number:\n\n```bash\nGLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088 \\\n  python3 packages\u002Fglyphsteer\u002Fexperiments\u002Femoji_collapse_check.py\n# mean off-diagonal cosine ≈ 1.000 ⇒ COLLAPSED (reject the model)\n# well below 1.0 (e.g. 0.77–0.97) ⇒ DISTINCT (proceed)\n```\n\nVerified: `bge-small-en-v1.5` → 1.000 (blind). `paraphrase-multilingual-MiniLM-L12-v2` →\ndistinct, with clean alignment. Fix a collapse with a one-env-var model swap + rebuild.\n\n## 3. Magnitude probe (Q1–Q3)\n\n```bash\nGLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088 \\\n  python3 packages\u002Fglyphsteer\u002Fexperiments\u002Fmagnitude_probe.py\n```\n- **Q1 (alignment):** `cos(glyph, aligned_word) − cos(glyph, unrelated)` — should be clearly\n  positive (the emoji means what you think). Measured +0.28…+0.58 with the multilingual model.\n- **Q2 (per-token nudge):** injecting one emoji shifts a neutral chunk's vector by Δcos\n  ~−0.03…+0.09 — real but modest, and a single token can wash out or slightly *invert* under\n  mean-pooling (e.g. ❓ measured −0.034). This is why Q2 is a tuning concern, not a guarantee.\n- **Q3 (anchor reshape):** a controllable gain (×2 at weight 1).\n\n## 4. Steer — the anchor metric reshape (pure numpy, host-side)\n\n`sim_eff` (lives in `glyphsteer.steer`, re-exported at top level — `from glyphsteer import\nsim_eff`) bends the similarity metric toward chosen anchors at query time:\n`sim_eff(q,d) = cos(q,d) + Σ_k w_k·⟨q,â_k⟩·⟨d,â_k⟩`, with â_k from `dense.measure_anchor`.\n\n```python\nfrom glyphsteer import dense, sim_eff\nanchor = dense.measure_anchor(\"😡\")     # the glyph's learned direction (via the sidecar)\nq = dense.embed([\"the deploy broke\"])[0]\ndocs = dense.embed([\"it crashed badly\", \"it shipped fine\"])\nscored = sim_eff(q, docs, anchors=anchor[None, :], weights=[1.5])   # weight = the steering knob\n```\nweight 0 = plain cosine; raising it makes that axis dominate (too high → ignores the query).\n\n## Honest status\nDense is **viable and measured** but its per-token strength is a *tuning knob*, not a free\nlunch; end-to-end dense retrieval lift on a real (non-toy) corpus is the open item. Lexical\nfaceting (the core **glyphsteer** skill) is the regime that's a clean win today.\n\n## Verify (run this)\n```bash\ncd packages\u002Fglyphsteer\u002Fserve && docker compose up --build -d && cd -\nexport GLYPHSTEER_EMBED_URL=http:\u002F\u002Flocalhost:8088\npython3 packages\u002Fglyphsteer\u002Fexperiments\u002Femoji_collapse_check.py   # gate\npython3 packages\u002Fglyphsteer\u002Fexperiments\u002Fmagnitude_probe.py        # Q1–Q3\n```\nArchitecture\u002Fflows: `packages\u002Fglyphsteer\u002F.claude\u002Frules\u002F{10,20}-architecture-*.md` (DW-6\u002FDW-7).\n",{"name":5,"description":7},"main","5117640090da92f427e1e21b431cfb53e7a3fb13",".claude\u002Fskills\u002Fglyphsteer-dense","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler","https:\u002F\u002Fgithub.com\u002Fsancovp","https:\u002F\u002Fisaacwr.carrd.co\u002F","AKA Olivus Victory-Promise\r\n\r\nAI Agents, Tools, and Infrastructure","2026-11-13 21:35:40",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add sancovp\u002Fchaincompiler --skill glyphsteer-dense",{"tool":38,"command":39},"bunx","bunx skills add sancovp\u002Fchaincompiler --skill glyphsteer-dense",{"tool":41,"command":42},"pnpm","pnpm dlx skills add sancovp\u002Fchaincompiler --skill glyphsteer-dense","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fglyphsteer-dense\u002FSKILL.md",true,[],[47,54,61,68,75,82],{"skillNo":48,"name":49,"nameZh":6,"description":50,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M75VCHMSPK8F1DNMR01SH2NJ","skill2framework","Turn an already-built AIOS\u002Fpackage into a framework chapter — narrative blog + deep-dive blog → chapter → {aios}-volume index skill → plugin → folded into the author's SkillTome.","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Fskill2framework\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M75VCDVAZBN85DC6V38A7Y8Z","package-plugin","WHAT: package a framework as a Claude Code PLUGIN (its operational face) — run skill2framework.package_plugin to assemble plugin.json + skills\u002F{aios}-volume\u002FSKILL.md + the chapter bundled under resources\u002F, with the skill's chapter links repathed to the bundle. WHEN: when a framework skill + chapter exist and need the installable plugin, or when the user says package the framework \u002F plugin \u002F skill2framework (any of).","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Fpackage-plugin\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M75VC9VRVDQ2PAT69P8PNB7G","narrative-blog","WHAT: produce Blog 1 of a framework CHAPTER — the narrative chapter-opener — for an already-built AIOS\u002Fpackage by reconstructing its journey from its own dir (CLAUDE.md, README, dev log) and FILLING skill2framework.JourneyCore, then rendering with render_blog1 (never hand-write the blog). WHEN: when making a framework\u002Fchapter from something already built, when you need the story post that opens a chapter, or when the user says blog organ \u002F narrative blog \u002F skill2framework (any of).","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Fnarrative-blog\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M75VC59SRRK5DZEWRZQV6SFW","framework-skill","WHAT: generate the FRAMEWORK SKILL for an already-built AIOS — the {aios}-volume (its SkillVolume face): one skill that indexes the AIOS's skills + rules + how they relate (as a diagram), gives a decision tree for what to do when, and ROUTES the agent into the AIOS dir. WHEN: when producing the index\u002Frouter face of a framework package, or when the user says framework skill \u002F volume \u002F skill2framework (any of).","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Fframework-skill\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M75VC0VS9AMREV5RMBVZ0RQK","fold-into-tome","WHAT: the TERMINAL stage — fold a finished framework into the author's SkillTome (the top-level holder + search of all their frameworks) by running skill2framework.fold_into_tome, which registers the framework as a row on a tome volume node's generated ## Frameworks table via skilltree's bind op. WHEN: when a framework package is finished and must be registered in the author's tome\u002Fvolume, or when the user says fold into the tome \u002F nomicon \u002F skill2framework fold (any of).","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Ffold-into-tome\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M75VBWYNPC80Y0HJC6WQKGQD","deepdive-blog","WHAT: produce Blog 2 of a framework CHAPTER — the mechanical deep-dive ('how it actually works') — for an already-built AIOS\u002Fpackage by reading its impl docs (README architecture sections, design docs, the code's module docstrings) and assembling a faithful mechanics-only article that links back to Blog 1. WHEN: when producing the deep-dive half of a chapter, when you need the technical how-it-works post Blog 1 links to, or when the user says blog 2 \u002F deep dive \u002F skill2framework (any of).","https:\u002F\u002Fgithub.com\u002Fsancovp\u002Fchaincompiler\u002Fblob\u002Fmain\u002Fskills\u002Fskill2framework\u002Fdeepdive-blog\u002FSKILL.md",[],[]]