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/aneduaim/untap-mcpWrote 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/aneduaim/untap-mcp/claude)<a href="https://agentmods.dev/plugins/aneduaim/untap-mcp/claude"><img src="https://agentmods.dev/badge/plugins/aneduaim/untap-mcp/claude/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/aneduaim/untap-mcp/claude"><img src="https://agentmods.dev/badge/plugins/aneduaim/untap-mcp/claude.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
untap 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 7d 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": "untap",
"displayName": "Untap",
"version": "1.3.3",
"description": "Untap checks UK refunds and compensation from delayed trains, disrupted flights, TfL journey charges and parking notices. Public checks return the applicable rules, amount, deadline and claim guidance. Sign in to save claims, track progress and check saved train commutes for dates you request. You submit claims directly to the operator or airline. Untap never files a claim for you.",
"author": {
"name": "Andrey Chekunin",
"email": "[email protected]",
"url": "https://untap.money/story"
},
"homepage": "https://untap.money/connect",
"repository": "https://github.com/aneduaim/untap-mcp",
"license": "MIT",
"keywords": [
"refunds",
"compensation",
"delay-repay",
"UK261",
"EU261",
"TfL",
"trains",
"flights"
],
"icon": "./assets/logo.png",
"documentationUrl": "https://untap.money/connect/docs",
"supportUrl": "https://untap.money/connect/docs",
"privacyPolicyUrl": "https://untap.money/privacy",
"termsOfServiceUrl": "https://untap.money/terms"
}
What it installs
The manifest is a name and a version. 1 MCP server travel with it, and installing the plugin installs all of 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.
- 7d ago Changed da18903f08bc
- 9d ago Changed fe30036de21d
- 10d ago Changed a198c6b0a659
- 11d ago First seen · 30 lines scan A 4152a21c706d
untap is a plugin published in the GitHub repository aneduaim/untap-mcp (0 stars, last pushed 8d 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 2026-10-05.
Other plugins, from other repositories
x402-list
Discover and vet x402 payment services before your agent pays one: uptime, price, and on-chain-verified settlement volume, from Gemini CLI.
drillr
drillr financial data and research MCP for AI agents — filings, statements, earnings, ownership, events, executives, analyst data, company discovery and research signals for US, China and Japan. Ten tools, browser sign-in, every figure traced to its filing.
drillr
drillr — financial data and research MCP for AI agents: filings with fact search, statements, earnings, ownership, events, executives, analyst data, company discovery and research signals for US, China and Japan.
optionsahoy-mcp
Deterministic equity-compensation tax tools: ISO/AMT exercise scheduling, NSO, RSU, QSBS, concentration risk, hedging, and equity funding plans.
horizon-shield
Check Japanese construction and renovation estimates against the open JCCDB dataset (526,128 records) and get Bitcoin-anchored, recomputable fair-price receipts; find contractors that passed the independent fair-price check (YAKUMO); public construction-cost data for Japan and the U.S. (JCCDB, USCCDB) from a separate…
redreplier
Monitor Reddit, Hacker News, X, Bluesky, and Facebook for keyword mentions, manage websites/keywords, triage AI-scored lead mentions, and configure alerts via the RedReplier API.