Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add shimo4228/claude-harness --skill jsonld-knowledge-graphgit clone --depth 1 https://github.com/shimo4228/claude-harnessWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/shimo4228/claude-harness/jsonld-knowledge-graph)<a href="https://agentmods.dev/skills/shimo4228/claude-harness/jsonld-knowledge-graph"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/jsonld-knowledge-graph/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/shimo4228/claude-harness/jsonld-knowledge-graph"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/jsonld-knowledge-graph.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00061 | $0.05930 |
| Opus 5.5 | $0.00024 | $0.02372 |
| Sonnet 5.5 | $0.00012 | $0.01186 |
| Haiku 4.5 | $0.00006 | $0.00593 |
Grade A, and why
jsonld-knowledge-graph scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
Copies of this mod
1 near-identical copy found in the catalogue:
- jsonld-knowledge-graph — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 314 lines — stays where its author put it; the contents beside it link to each section on GitHub.
JSON-LD Knowledge Graph
graph.jsonld を llms.txt の隣に置いて、project の concept-level architecture を schema.org JSON-LD triples として encode する skill。LLM が prose だけでは引き出しにくい entity 間の関係（matrix / hierarchy / phase-binding）を triple として citation 可能にする。
When to Use
以下を すべて 満たす project に適用する:
- 安定した concept-level 構造を持つ
- 例: matrix（4 Quadrants × N ADRs）、ordered hierarchy（prohibition-strength 3 levels）、phase-skill binding（6:6 bijective）、layered architecture（3 memory layers）
- 構造が release を跨いで安定（vX.Y.Z bump で entity が増減しない）
- LLM に当該関係を訊いた時に prose だけでは答が出にくい / 誤答する 経験的証拠がある
- すでに
llms.txt+llms-full.txt（Answer.AI 標準）を持っている
When NOT to Use
以下のいずれかに該当する場合は 適用しない:
- 単一目的の linear project（matrix / hierarchy 構造がない）
- 内部構造が release ごとに churn する（graph が編集負債になる）
llms-full.txtの prose Q&A で関係問題が既に十分答えられる- 「LLM citation 向上を狙いたい」だけで具体的な entity 関係問題がない（GEO 目的なら llms-txt-writer の方が direct）
graph.jsonld は prose の代替ではなく 補完。prose で表現しきれない構造があるとき、その構造をやっと拾える。
file-level 構造との関係（正本）
graph.jsonld は concept 層だけ を保存する。file-level 構造（「X はどのファイルに住むか」「誰が誰を呼ぶか」）は保存せず、問いのたびにコードから導出する:
| file-level（導出、保存しない） | graph.jsonld（保存） | |
|---|---|---|
| 対象 | ファイル / モジュール / call 関係 | 概念 / エンティティ |
| 答える質問 | 「X はどのファイルに住んでいるか」「誰が Y を呼ぶか」 | 「X とは何か、X と Y はどう関係するか」 |
| 源 | コードそのもの — Claude Code の LSP tool（workspaceSymbol / findReferences / incomingCalls）、grimp 等の import グラフ |
JSON-LD triples（手で書く） |
| 主読者 | 作業中の agent | AI search engine + LLM が entity を citation する時 |
| trigger | なし（都度計算） | concept / 関係の変化 |
graph にコードのノード（file path・module・LOC）を置かない。置くと第 2 の module map になり、ソース commit ごとに同期コストを払う鏡が育つ（手書き module map 6 枚がソース 197 commit に対し 159 commit の同期を要し、読者は観測されなかった実例がある）。設計理由は ADR、パイプラインの段構成はそれを走らせる script の冒頭コメントが持つ。
Drift 防止のための運用規約
- 新規 ADR / Concept / Quadrant 等を追加 → graph.jsonld にノード追加（
@idは GitHub blob URL） - ファイルパスが変わった → graph.jsonld の
@idが blob URL ならそこだけ追従。他に更新する文書は無い - 概念の semantics が変わった → graph.jsonld の description 更新
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago Changed · +2 lines 2b8397092f5d
- 11d ago Changed · -7 tokens per session 45cc4b91e8b9
- 13d ago Changed · -12 lines · -22 tokens per session 427102108fa6
- 25d ago Changed a5ad0992f875
- 1mo ago Changed · -62 lines 94f4204d997f
- 1mo ago Changed · -7 lines e16a8573b8c2
- 1mo ago Changed c707dc4582b8
- 1mo ago First seen · 393 lines · 90 tokens per session scan A a95604e01a49
jsonld-knowledge-graph is a skill published in the GitHub repository shimo4228/claude-harness (3 stars, last pushed 2d ago), licensed MIT. It adds 61 tokens to every session and 5,930 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
inference-engineer
Productize an open-source model into a hosted inference endpoint the researcher (or their agent) can call. Picks the right hardware, the right serving stack (vLLM / Triton / TEI / BentoML), wraps it in an OpenAI-compatible gateway (LiteLLM) with per-tenant auth, exposes it as an MCP tool in chat, and runs a quality +…
experiment
Run a materials-science / ML compute job on Rockie GPU capacity. Trigger words "run experiment", "submit job", "/experiment", or requests to quote/approve GPU spend before an experiment. Picks the right GPU type and count from a natural-language description (DFT for QE/VASP/ABINIT, MD for GROMACS/LAMMPS/OpenMM…
finetune-model
Track 2 quickstart wrapper for structured fine-tune prompts using a registry model, a registry dataset, Rockie GPU jobs, and inference-loader deployment of the trained artifact.
sota-delta
Track 3 quickstart wrapper that reproduces a paper or repository baseline on Rockie GPU, then extends it with a user-specified delta and compares baseline versus delta.
deploy-model
Narrow quickstart wrapper for pre-baked Deploy-card requests. Triggers on server-validated structured quickstart deploy prompts such as "Quickstart deploy request: track: deploy model: source: quickstart-picker quickstartintentid: " and explicit "deploy this featured model" requests from Rockie's quickstart picker…
autopilot
Continuous-operation mode for rockie — runs the experiment queue autonomously, using Zero-Cost Monitoring ($0 LLM cost during training), anti-burn exponential cooldown on failures, and ntfy to wake the human only when a decision is needed. Use when you want agent-driven research to proceed for days without human…