I built an efficient graph-search plugin for Claude Code skills

Developer Daniel Lublinsky released Skill Atlas, a free, open-source graph-search plugin for Claude Code that adds a 'searchable' tier between enabled and disabled skills, cutting per-session token usage from 2,544 to 338 tokens (−86.7%) across 53 skills by keeping only 6 enabled. The plugin, installable via 'claude plugin marketplace add danielLublinsky/Skill_Atlas', builds a graph index and renders an HTML atlas, with a search costing about 2.4k tokens only when invoked.

A third tier for Claude Code skills — dormant, zero tokens, still findable. Search a graph of your collection instead of preloading it. 53 skills, 6 kept enabled: 2,544 → 338 tokens per session −86.7% . The other 47 cost nothing until asked for. claude plugin marketplace add danielLublinsky/Skill Atlas claude plugin install skill-atlas@skill-atlas Python 3 only — no dependencies, no network d3 is vendored . Later: claude plugin update skill-atlas@skill-atlas . From a local clone for hacking on it git clone https://github.com/danielLublinsky/Skill Atlas.git claude plugin marketplace add ./Skill Atlas claude plugin install skill-atlas@skill-atlas | Command | What it does | |---|---| /skill-atlas | Build the graph, categorize what's new, render atlas.html . Run this once to set up. | /skill-atlas:edit-searchable | Move plugins in and out of the searchable tier, interactively. | /skill-atlas:skill-search | Find the right skill for a task across the whole library, dormant tier included. Usually you never type this — Claude calls it on its own before a nontrivial task. | automatic | A SessionStart hook keeps everything fresh after that. | Then open ./.claude/skill-atlas/atlas.html in a browser, and just ask for a task — the search runs itself, names the skill it picked in one line, and gets on with the work. Without the plugin Makefile make build graph.json + catalog/ from your live manifests make render atlas.html make check CI gate: exit 1 on any broken reference or dangling mention make test unit suite against fixtures — never touches the real ~/.claude Claude Code injects every enabled skill's description into every session — ~48 tokens each. At 100+ skills that's thousands of tokens you pay for constantly, and near-duplicates quietly compete for the model's attention. Disabling fixes the cost and loses the skill. So Skill Atlas adds a tier in between: | Tier | In context? | Findable? | Cost / session | |---|---|---|---| 🟢 enabled | yes | natively | ~48 tokens each | 🟣 searchable | no | via skill-search | 0 | ⚫ disabled | no | no | 0 | A searchable skill is dormant — its description never enters your context, but it stays discoverable on demand. All that advertises the whole dormant tier is one ~50-token line at session start. flowchart LR A "🧭 a task arrives<br/ <i “resolve this merge conflict”</i " subgraph R1 "1️⃣ read the index · ~1.6k tok" I "<b index.md</b — 14 lines<br/ <i name · count · what it covers · member tokens</i " end subgraph R2 "2️⃣ read one shard · ~0.6k tok" S "<b version-control.md</b <br/ using-git-worktrees<br/ resolving-merge-conflicts ← hit" end X "the other 13 shards<br/ <i ~10k tok · never opened</i " A -- I I -- "pick ONE category" -- S I -. skipped .- X S -- E "🟢 <b enabled</b <br/ <i invoke natively</i " S -- F "🟣 <b searchable</b <br/ <i read its path,<br/ follow as instructions</i " classDef file stroke: 8a8578,stroke-width:1.5px classDef ghost fill:transparent,stroke: 8a8578,stroke-width:1.5px classDef hit fill: 1baf7a26,stroke: 1baf7a,stroke-width:2px classDef dormant fill: 8a5cd626,stroke: 8a5cd6,stroke-width:2px classDef muted fill:transparent,stroke: 8a8578,stroke-width:1px,stroke-dasharray:4 3 class I,S file class A ghost class E hit class F dormant class X muted style R1 fill: 3987e514,stroke: 3987e5,stroke-width:1.5px style R2 fill: 3987e514,stroke: 3987e5,stroke-width:1.5px A search costs ~2.4k tokens — the index 1.6k , one shard ~0.6k and the search skill itself 0.4k — and you pay it only when a search happens. The alternative it replaces is every dormant description sitting in context from session start, billed whether you search or not. The arbitrage is real when searches are occasional; it is not free, and the numbers above are measured rather than estimated. One shard is the norm. A second is opened only when the entry you found lists a category you haven't read — the catalog pointing, not the model guessing — and never a third. Counts overlap on purpose, so the pick doesn't have to land on the right bucket, only a right one. Tier-off skills are excluded when shards are written, so a disabled plugin can never return through a side door. The result is announced in one line before the work continues: search is a lookup inside a task, not a deliverable. The model categorizes once. The first run drafts 8–12 categories named for user-intent task shapes , not products, then freezes the taxonomy. Later runs only file new skills and re-confirm the ones whose description changed — each assignment carries a hash of the description it was made against. Discovery is manifest-driven — installed plugins.json → each plugin.json → settings — so marketplace catalogues and stale cached versions never pollute the picture. The naive "every directory with a SKILL.md" count over-counts by ~60% on a real machine: build graph.py --naive-count . Skills are nodes. The edges are the interesting part: | Edge | Caught | | |---|---|---| | 🔗 | references — a skill → its own bundled files | broken when the file isn't there | | 💬 | mentions — a skill naming another skill | dangling when the target is disabled , unregistered or absent | Mention matching is deliberately strict — only backticked, skills/<name , or unambiguous hyphenated names, code fences stripped first. Loose matching produced 91 edges on a 41-skill collection, almost all of them ordinary English words. You also get duplicate names, orphans, and an exit code to gate CI on. atlas.html is self-contained and opens from file:// with zero network requests: scope origin, state, breakage in red and categories hub-and-spoke — solid edge to the home category, dashed to the rest; search cat:<name to isolate . What's really registered — discovery is manifest-driven, so unregistered copies and stale plugin caches surface as their own nodes instead of quietly counting. Broken bundles — a references edge whose file isn't on disk draws red, so a typo'd references/foo.md is visible without opening anything. Why it didn't trigger — tier is the node's fill enabled · searchable · disabled , and duplicate names are called out in the footer with both nodes drawn. Read it in place — pin any skill or bundled file and hit open markdown to read the source in a popup without leaving the graph. docs/ /danielLublinsky/Skill Atlas/blob/main/docs — component guide, numbered 1–9: start at overview /danielLublinsky/Skill Atlas/blob/main/docs/1-overview.md , then jump to the component you're touching discovery, graph build, categorization, catalog & search, rendering, hooks . DESIGN.md /danielLublinsky/Skill Atlas/blob/main/DESIGN.md — the historical record for both phases: what was considered, what was chosen, what was dropped, and why Apache-2.0 — see LICENSE /danielLublinsky/Skill Atlas/blob/main/LICENSE . D3 is vendored in vendor/d3.v7.min.js /danielLublinsky/Skill Atlas/blob/main/vendor/d3.v7.min.js and inlined into every generated atlas.html . It is ISC-licensed, © 2010-2023 Mike Bostock — see vendor/LICENSE-d3 /danielLublinsky/Skill Atlas/blob/main/vendor/LICENSE-d3 .