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.
git clone --depth 1 https://github.com/davisjam/model-based-agentic-software-engineeringWrote 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/plugins/davisjam/model-based-agentic-software-engineering/mage)<a href="https://agentmods.dev/plugins/davisjam/model-based-agentic-software-engineering/mage"><img src="https://agentmods.dev/badge/plugins/davisjam/model-based-agentic-software-engineering/mage/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/plugins/davisjam/model-based-agentic-software-engineering/mage"><img src="https://agentmods.dev/badge/plugins/davisjam/model-based-agentic-software-engineering/mage.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
mage 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 1mo 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.
What it actually says
{
"name": "mage",
"description": "The mage toolkit — the Davis AI-First Engineering Method plus the control catalogue, packaged as installable skills. Ships three partner skills: self-governance (harden — audit a repo for missing guardrails, or convert a recurring failure into a durable control), self-operations (operate — run the agent-fleet substrate; route a break to its fix, and escalate a recurring one to self-governance), and self-communicate (communicate — write and diagram engineer-facing docs well: rhetoric, the Diátaxis register, a house lexicon, and a prose+diagram audit).",
"version": "0.1.0",
"license": "MIT",
"author": {
"name": "James C. Davis, PhD",
"email": "[email protected]"
},
"homepage": "https://davisjam.github.io/agent-governance-mechanisms/",
"repository": "https://github.com/davisjam/agent-governance-mechanisms",
"keywords": [
"governance",
"agents",
"ai-engineering",
"linting",
"code-review",
"guardrails",
"mbse"
]
}
What it installs
The manifest is a name and a version. 3 skills travel with it, and installing the plugin installs all of them — 773 tokens a session between them. Each is measured on its own page, and each can be installed alone.
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.
- 1mo ago First seen · 22 lines scan A 810b84d4f2d0
mage is a plugin published in the GitHub repository davisjam/model-based-agentic-software-engineering (25 stars, last pushed yesterday), 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 2026-08-30.
Other plugins, from other repositories
dx
Developer experience essentials: GitHub Actions debugging, conversation half/quarter-cloning, context handoffs, research (Reddit and Hacker News), private GitHub search, and Claude Code version checks.
credential-brokers
Your credentials stay out of the repo and out of the agent's context. credbroker resolves them in-process, sso-broker handles the browser cookie dance, and an LLM-cooperative credential-setup skill walks you through the first time.
lanes-bridges
Bridge Lanes with GitHub and Linear: import issues, batch-spawn a session per ticket, and post results back upstream. Split out of lanes-desktop in 0.7.0.
bestpractice
Enforcement, memory and parallel-session coordination for solo founders running several Claude Code sessions on one repository.
agent-evolutions
Genetic exploration of solution spaces for AI coding agents — gather a verifiable objective, evolve variants, apply the winner.
ooda-loop marketplace
Plugin marketplace listing 1 plugin: ooda-loop.