{"id":"2023957499183829467","url":"https://x.com/arscontexta/status/2023957499183829467","text":"","author":{"name":"Heinrich","username":"arscontexta","avatarUrl":"https://pbs.twimg.com/profile_images/2012958446891536384/neq1Tu46_200x200.jpg"},"createdAt":"Wed Feb 18 03:07:25 +0000 2026","engagement":{"replies":208,"retweets":903,"likes":8823,"views":4067117},"article":{"title":"Skill Graphs > SKILL.md","previewText":"people underestimate the power of structured knowledge. it enables entirely new kinds of applications\nright now people write skills that capture one aspect of something. a skill for summarizing, a","coverImageUrl":"https://pbs.twimg.com/media/HBaFlJaW4AAfpA8.jpg","content":"people underestimate the power of structured knowledge. it enables entirely new kinds of applications\n\nright now people write skills that capture one aspect of something. a skill for summarizing, a skill for code review and so on. (often) one file with one capability\n\nthats fine for simple tasks but real depth requires something else\n\nimagine a therapy skill that provides relevant information about cognitive behavioral patterns, attachment theory, active listening techniques, emotional regulation frameworks and so on\n\na single skill file cant hold that\n\n## skill graphs\n\na skill graph is a network of skill files connected with wikilinks\n\ninstead of one big file you have many small composable pieces that reference each other. each file is one complete thought, technique or skill and [[wikilinks between them create a traversable graph]]\n\na skill graph applies the same skill discovery pattern recursively inside the graph itself\n\nevery node has a yaml description the agent can scan without reading the whole file\n\nevery wiki link carries meaning because its woven into prose so the agent follows relevant paths and skips what doesnt matter\n\nprogressive disclosure:\n\n> index → descriptions → links → sections → full content\n\nmost decisions happen before reading a single full file\n\n## the primitives\n\nyou already have everything you need\n\n- wikilinks that read as prose in sentences, so they carry meaning not just references\n\n- yaml frontmatter with descriptions so the agent can scan without reading full files\n\n- MOCs (maps of content) that organize clusters of related skills into navigable sub-topics\n\nskill links to other skills which link to other skills and the graph goes as deep as the domain requires\n\n## arscontexta plugin\n\narscontexta is a skill graph that teaches your agent how to build skill graphs\n\n(okay actually its about building knowledge bases but thats the same thing...)\n\n~250 connected markdown files that teach an agent how to build a massive knowledge base aka skill graph for you\n\none skill file couldnt do that\n\nbut things change if you build a graph of interconnected research claims (/skills) about cognitive science, zettelkasten, graph theory, agent architecture where each piece links to others, each one composable and the whole thing is traversable\n\n## what this enables\n\nthink about it:\n\n- a trading skill graph: risk management, market psychology, position sizing, technical analysis, each piece linked to related concepts so context flows between them\n\n- a legal skill graph: contract patterns, compliance requirements, jurisdiction specifics, precedent chains, all traversable from one entry point\n\n- a company skill graph: org structure, product knowledge, processes, onboarding context, culture, competitive landscape\n\nnone of these fit in one file but all of them work as graphs\n\n## how to build one\n\nthe easy way: install the arscontexta claude code plugin, pick the research preset and point it at any topic\n\nit sets up the markdown folder structure for you and then you fill it with /learn and /reduce\n\nthe manual way its simpler than you think\n\na skill graph doesnt need to live in your .claude/skills/ folder. the key is an index file that tells the agent what exists and how to traverse it\n\nheres what an index looks like for a knowledge work skill graph:\n\nthe index isnt a lookup table its an entry point that points attention. the agent reads it, understands the landscape and follows the links that matter for the current conversation\n\neach linked file is a standalone methodology claim (= skill). heres what one node looks like:\n\n![](https://pbs.twimg.com/media/HBaKDqcXoAAT1ds.jpg)\n\nsee how the wikilinks inside the prose tell the agent when and why to follow them\n\nan map of contents (MOCs) organize sub-topics when the graph gets larger.\n\n## the evolution\n\nskills are context engineering, basically curated knowledge injected where it matters\n\nskill graphs are the next step\n\ninstead of one injection the agent navigates a knowledge structure, pulling in exactly what the current situation requires\n\nthis is the difference between an agent that follows instructions and an agent that understands a domain\n\narscontexta is a claude code plugin that does this for building knowledge systems. 249 files of structured knowledge the agent traverses to derive a local knowledge system that really fits your workflow\n\ngo use it and build skill graphs for everything else\n\nheinrich"},"adhxContext":{"savedByCount":1,"publicTags":[],"previewUrl":"https://adhx.com/arscontexta/status/2023957499183829467"}}