Borrowing it
Nothing to install: this file belongs to zenml-io/kitaru. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/zenml-io/kitaru/develop/.agents/skills/kitaru-adapter-development/SKILL.mdgit clone --depth 1 https://github.com/zenml-io/kitaruWrote 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/zenml-io/kitaru/kitaru-adapter-development)<a href="https://agentmods.dev/skills/zenml-io/kitaru/kitaru-adapter-development"><img src="https://agentmods.dev/badge/skills/zenml-io/kitaru/kitaru-adapter-development/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/zenml-io/kitaru/kitaru-adapter-development"><img src="https://agentmods.dev/badge/skills/zenml-io/kitaru/kitaru-adapter-development.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00051 | $0.01115 |
| Opus 5 | $0.00026 | $0.00558 |
| Sonnet 5 | $0.00010 | $0.00223 |
| Haiku 4.5 | $0.00005 | $0.00112 |
Grade A, and why
kitaru-adapter-development 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Kitaru Adapter Development
Read AGENTS.md. For Python adapters, also read plugins/AGENTS.md and plugins/DEVELOPMENT.md. For TypeScript adapters, read release/typescript.md before changing package or release metadata. Load the same-name kitaru-dev repo skill for the current host for the general command and PR workflow.
Start from the closest current adapter and the target framework's public API. Do not port a historical adapter wholesale.
Choose the extension point
Recording adapters are independent Python distributions under plugins/packages/<slug>/, with focused tests under plugins/tests/adapters/<slug>/. Importer-backed adapters are different: they live inside the provider's importer package as an ImporterBackedAdapter subclass, never intercept model or tool calls, and follow the kitaru-importer-development skill instead of this section's package rules. Agent versions running them declare runtime_capabilities with overrides: false and tool_policies: false on the run spec. TypeScript framework adapters are packages such as packages/mastra/ and packages/vercel-ai/; shared adapter primitives live under packages/core/src/adapter/.
Before editing, state:
- which public framework call, hook, callback, or wrapper is intercepted
- which model, tool, handoff, child-agent, and failure events are observable there
- when sessions and nodes are written, including what the caller observes if recording fails
- which replay overrides can be applied at a real framework boundary
- which streaming, tool, approval, resume, stateful, or dynamic cases remain unsupported
Preserve the framework's ordinary behavior, configured hooks and state, public entrypoint, and native result type. Reject unsupported replay configurations before model or tool execution. Treat passthrough tools as real side effects, not as a reversible transaction. If the framework has no public per-run model replacement point, do not emulate one by mutating shared agent state or reading private fields; stop and report that replay boundary.
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.
- 4d ago Changed · +2 lines 64160ec49a50
- 5d ago Changed cd6b0771f385
- 9d ago First seen · 64 lines · 51 tokens per session scan A 6e0db29b5d6c
kitaru-adapter-development is a skill published in the GitHub repository zenml-io/kitaru (282 stars, last pushed today), licensed Apache-2.0. It adds 51 tokens to every session and 1,115 once invoked, about $0.0003 per session on Opus 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 2026-08-30.
Other skills, from other repositories
agent-lightning
Provides the action space, tradeoffs, and evaluation context for improving an editable AI agent against a benchmark while preserving its deployment contract. Use when optimizing agent accuracy, cost, latency, or reliability.
huggingface-hub
HuggingFace hf CLI: search/download/upload models, datasets.
healthchain
Use when building, debugging, or deploying a Python service that touches FHIR resources, EHR APIs, CDS Hooks, clinical documents, or patient data — including writing model or agent output back into a patient record, connecting to Epic/Cerner/Medplum, or serving FHIR tools to an agent over MCP or LangChain.
ai-architect-expert
Expert-level AI system design, MLOps, architecture patterns, and AI infrastructure. Use when the user mentions AI architecture, MLOps, system design, AI infrastructure, or scalability, or when the task involves AI System Architecture, MLOps Infrastructure, Scalability Patterns, or Architecture Design.
ml-expert
Expert-level machine learning, deep learning, model training, and MLOps. Use when the user mentions machine learning, deep learning, neural networks, MLOps, or data science, or when the task involves Machine Learning Fundamentals, Data Preparation, or Model Training.
headless
Autonomer Langlauf-Modus ohne laufende Rückfragen — Startbedingungen (Ziel, Scope, Budgets, Checkpoints), Entscheidungsstufen und Abbruchregeln für unbeaufsichtigte Läufe; erhöht die Ausdauer, nicht die Berechtigung. Nutzen, wenn ausdrücklich ein längerer autonomer Lauf gewünscht ist.