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 skills add aneduaim/untap-mcp --skill find-travel-refunds-in-inboxgit 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/skills/aneduaim/untap-mcp/find-travel-refunds-in-inbox)<a href="https://agentmods.dev/skills/aneduaim/untap-mcp/find-travel-refunds-in-inbox"><img src="https://agentmods.dev/badge/skills/aneduaim/untap-mcp/find-travel-refunds-in-inbox/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/aneduaim/untap-mcp/find-travel-refunds-in-inbox"><img src="https://agentmods.dev/badge/skills/aneduaim/untap-mcp/find-travel-refunds-in-inbox.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00053 | $0.00840 |
| Opus 5.5 | $0.00021 | $0.00336 |
| Sonnet 5.5 | $0.00011 | $0.00168 |
| Haiku 4.5 | $0.00005 | $0.00084 |
Grade A, and why
find-travel-refunds-in-inbox 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 9d 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 — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Find travel refunds in an inbox
Use this skill when you can read the person's email and they ask you to find money they are owed ("find my refunds", "am I owed anything", "check my emails for compensation"), or during a periodic sweep they have asked for. It finds candidate train, flight and TfL claims and hands them to Untap (https://untap.money), which checks them against real records.
Searching is read-only. Submitting anything follows the uk-travel-claims skill and needs the person's say-so.
Steps
- Search narrowly. Run the queries in references/email-signals.md: operator, retailer, airline and TfL senders, plus disruption wording in subjects. Use the windows below. Do not crawl the whole mailbox.
- Skim, then open. Read subjects and dates first. Open only booking confirmations and disruption notices. Skip the false positives listed in the reference (marketing, price alerts, future trips, refunds already paid).
- Pair and group. Match each disruption email with its booking (same reference, date, route). One journey is one candidate.
- Hand to Untap. Send the travel emails to
scan_for_claims(up to 25 per call:from,subject,date,body_text, trimmed of unrelated text). Run the check tool it suggests with its prefilled arguments:check_train_delay,check_flight_compensationorcheck_tfl_journeys. Ifscan_for_claimsis not available, extract the fields yourself and call the check tool directly. - Report. Show a short list, nearest deadline first: the journey, what happened, Untap's amount, the deadline, and what is still missing (often the fare or which train). Say which emails you could not resolve and why.
- Stop and ask. Ask which claims the person wants to make. Then follow the
uk-travel-claimsworkflow: confirm details, fill in the official form with permission, record, chase.
| Category | Look back | Why |
|---|---|---|
| Trains | 35 days (60 to pair bookings) | Delay Repay closes 28 days after travel |
| TfL | 56 days | Incomplete journey refunds close after 8 weeks |
| Flights | 12 months first, older on request | Untap's flight records cover about a year; claims can run 6 years (5 in Scotland) |
What ships with it
1 file beside SKILL.md 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.
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.
- 9d ago Changed 834f3fda78aa
- 10d ago First seen · 44 lines · 53 tokens per session scan A fc086232365e
find-travel-refunds-in-inbox is a skill published in the GitHub repository aneduaim/untap-mcp (0 stars, last pushed 6d ago), licensed MIT. It adds 53 tokens to every session and 840 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-10-05.
Other skills, from other repositories
x402-list
Discover x402 payment services and on-chain-verified facilitator settlement volume via the x402-list directory. Use when a user or agent needs to find an API/service that accepts x402 (HTTP 402 stablecoin) payments, compare services by reliability and price, check service uptime/health, see whether a service has…
redreplier
Monitor Reddit, Hacker News, X, Bluesky, and Facebook for keyword mentions of your product/website via the RedReplier API. Covers managing monitored websites, keyword lifecycle (add/edit/disable/enable/delete/activate within the plan), triaging AI-scored lead mentions (approve/reject, relevance reasoning), and email…
plan-equity-portfolio
Build one prioritized plan across a person's whole equity position - incentive stock options (ISOs), non-qualified stock options (NSOs), restricted stock units (RSUs), and held company stock - instead of answering one grant at a time. Use when someone has more than one grant type, asks "what should I do with my…
equity-comp-tax
Equity-compensation tax planning. Use when the user asks about exercising incentive stock options (ISO) or the alternative minimum tax (AMT), non-qualified stock option (NSO) exercise tax, restricted stock unit (RSU) vesting decisions, Section 1202 Qualified Small Business Stock (QSBS) qualification, single-stock…
adaptlypost
Create, schedule, and manage social media posts across Instagram, TikTok, YouTube, X, LinkedIn, Facebook, Pinterest, Threads, Bluesky, Mastodon, and Google Business Profile via the AdaptlyPost API, and read how they performed. Covers post creation, scheduling, recurring posts, bulk scheduling, per-platform results…
flowsery
Query web analytics data from Flowsery Analytics, a privacy-first web analytics platform. Retrieve real-time visitor counts, time series data, breakdowns by 25 dimensions (device, page, country, referrer, campaign, channel, UTM params, exit links, etc.), visitor profiles with activity timelines, custom goal tracking…