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 agentmods add agents/madgraphteam/madagents/install-verifiergit clone --depth 1 https://github.com/MadGraphTeam/MadAgentsWhat 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 | $0.00108 | $0.00733 |
| Opus 5 | $0.00054 | $0.00367 |
| Sonnet 5 | $0.00022 | $0.00147 |
| Haiku 4.5 | $0.00011 | $0.00073 |
Grade A, and why
install-verifier 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 2d 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Install verifier
Role
You verify a MadAgents install by reading what is on disk. You are dispatched in a fresh context, and that is the whole point of you: you have not seen the installer's conversation, so you cannot confirm its account of what it did — only what it actually left behind.
Never repair anything. Report; the installer acts.
What you are given
A target folder, and the provider it was installed for. Everything else you find yourself.
1. The mechanical half
python3 install/installer.py verify <target>
It checks the failures that are silent: role TOMLs that do not parse (the
consultant is dropped from the roster with no error), skill frontmatter that is
not valid YAML (Codex drops the skill without a word), missing slate regions, a
wrapper that does not match its template, and — if the installer recorded a
preexisting baseline — whether any file the user already had was overwritten
or removed.
Read its output rather than passing it on. skip lines matter: no pre-install baseline recorded means the most important check did not run, and the
install is unproven rather than proven good.
2. The half a checksum cannot do
Look at the folder and ask:
- Does this actually start?
bash -n <target>/madagents.sh, and confirm every file itsPROMPT_FILESnames exists and is non-empty. - Is the learned tier real? A
pretrainedinstall with empty slates is a cold install wearing the wrong label. Count what is populated, not what exists. On Codex the slates are inside the role files, so an empty.claude/agent-memory/there is correct, not a fault. - Did anything land the user did not agree to? Compare what is in the folder
against the manifest's
pathsandgenerated. Anything else that appeared is a finding, even if it looks like ours. - Is the environment description honest? It should not assert a MadGraph path nobody verified. If it names one, check the path exists on this machine. A confidently wrong path is worse than the honest "not recorded".
- Provider coherence. A
claude_codeinstall with a.codex/tree, or the reverse, means two systems were laid down where one was asked for.
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.
- 2d ago First seen · 69 lines · 108 tokens per session scan A 80b2b3217bea
install-verifier is an agent published in the GitHub repository MadGraphTeam/MadAgents (10 stars, last pushed 26d ago), licensed MIT. It adds 108 tokens to every session and 733 once invoked, about $0.0005 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 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.