A collection of developer guidance for typed language-model functions, Jinja2 templates, GNU Make, and Hugging Face model configuration and tokenizers. Jinja2 is a template language, while Make automates command sequences and build tasks.

Use it to write BAML functions, maintain Jinja2 templates for chat or Ansible, structure Makefiles, and check Hugging Face configurations before connecting them to vLLM or SGLang.

It helps prevent mismatches between templates, model settings, tokenizers, and the programs that serve models. It also provides established patterns for writing Makefiles and typed LLM functions.

dev — Plugin for Claude Code by air-gapped

Install

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.

Clone the repo

git clone --depth 1 https://github.com/air-gapped/skills

Made for: Claude Code.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/plugins/air-gapped/skills/dev.svg)](https://agentmods.dev/plugins/air-gapped/skills/dev)

Your own site

<a href="https://agentmods.dev/plugins/air-gapped/skills/dev"><img src="https://agentmods.dev/badge/plugins/air-gapped/skills/dev.svg" alt="Measured on agentmods" height="20"></a>

Per session not measured What this adds to a session before it is invoked.

When invoked not measured The manifest loads nothing itself; its 4 skills cost 631 tokens a session between them.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Security

Grade A, and why

dev 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 5d 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.

.claude-plugin/marketplace.json#dev · 31 lines

What it actually says

{
  "name": "dev",
  "source": "./",
  "description": "Developer tooling suite — BAML (typed LLM functions), Jinja2 templates (HuggingFace chat templates, Ansible), GNU Make best practices, and HuggingFace transformers config/tokenizer preflight (vLLM/sglang engine bridge).",
  "version": "0.20260826.23",
  "author": {
    "name": "Jörgen"
  },
  "license": "MIT",
  "skills": [
    "./.claude/skills/baml-expert",
    "./.claude/skills/jinja-expert",
    "./.claude/skills/makefile-best-practices",
    "./.claude/skills/transformers-config-tokenizers-expert"
  ],
  "strict": false,
  "category": "developer-tools",
  "tags": [
    "baml",
    "jinja",
    "makefile",
    "gnu-make",
    "chat-templates",
    "huggingface",
    "ansible",
    "typed-prompts",
    "transformers",
    "tokenizers",
    "preflight"
  ]
}

Contents

What it installs

The manifest is a name and a version. 4 skills travel with it, and installing the plugin installs all of them — 631 tokens a session between them. Each is measured on its own page, and each can be installed alone.

Files

What ships with it

1 file beside marketplace.json#dev 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.

marketplace.json 31 KB

Changes

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.

5d ago First seen · 31 lines scan A 2b0a6781d262

Subscribe to this mod's changes

dev is a plugin published in the GitHub repository air-gapped/skills (5 stars, last pushed 6d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. 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 2026-09-03.

Claude Code Data and AI DevOps and cloud Prompting agent-skills ai-agents air-gapped anthropic claude-code-marketplace

minimax

MiniMax M-series production wiring: OpenAI-compat quirks, hybrid prompt caching, Tier F quant agentic stack, model-upgrade detection. Distilled from a 41-iteration M2.7-highspeed exploration campaign.

terrylica/cc-skills · not measured

prompt-coach

Analyzes each prompt you submit against a tiered ruleset of prompting best practices; nudges you toward better prompts (definition-of-done, scoped references, guardrails, verification steps) and quietly graduates rules you consistently follow.

alexmond/alexmskills · not measured

prompt-triwizard

LLM prompt engineering for agent pipelines: write, refine, and evaluate in Anthropic, Google, or OpenAI style.

HiMyNameIsDavidKim/prompt-triwizard-skill · not measured

ai-toolkit

Skills for Deep Learning (Keras 3), Reinforcement Learning (Python & C++), Claude Prompt Optimization, x402 Web3 Payments, and OpenClaw Setup.

Aznatkoiny/zAI-Skills · not measured

ai-ml-skills

MLOps patterns, prompt engineering, LLM evaluation, and embedding optimization.

ancoleman/ai-design-components · not measured

llm-language

Self-evolving prompt meta-compiler for the Fable 5 / Opus 4.8 era (Mythos-class aware) with xhigh/max effort, adaptive thinking, task budgets, dual-layer memory (ROSETTA.md + auto-memory), AND skill-level meta-refinement. v4.4: FABLE-5-ERA RETARGET — complexity map gains a model dimension (Opus 4.8 workhorse default…

Silence-view/llm-language · not measured