Guide · Agent skill
How to use the Stablecoin Flow Analysis agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The stablecoin-flow skill is a set of written instructions that teaches an AI agent how to interpret stablecoin data, such as USDT and USDC supply, minting and burning, reserves on exchanges and on-chain transfer activity, as a rough sign of money moving into or out of crypto markets. This guide is for readers who have not used agent skills before. By the end you will know what the skill changes about the agent's answers, how to install it, how to phrase requests that trigger it, and what its limits are.
Stablecoin Flow AnalysisHKUDS/Vibe-Trading · MIT Download the skill Quick start
- 1Download the ZIP. Get the stablecoin-flow ZIP from this page and unzip it so you have a folder containing SKILL.md.
- 2Place the folder. Move it to ~/.claude/skills/ for personal use, or to .claude/skills/ inside a project.
- 3Or upload it in the apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the skill instead.
- 4Gather your data. Collect current stablecoin supply, mint and burn, exchange reserve and transfer figures from your own source, since the skill does not fetch them.
- 5Ask a specific question. Request an analysis in plain language and include the figures and a date range.
- 6Check the thresholds. Ask the agent which rules from the file were triggered and how the composite score was built.
- 7Use it as context only. Review the result critically, because the signals are heuristic and not validated for trading decisions.
What the stablecoin-flow skill makes the agent do
Once installed, the agent frames its answers around stablecoin indicators. The skill's file covers five areas: total stablecoin supply and its 30-day change, mint and burn events for USDT and USDC, exchange stablecoin reserves, stablecoin dominance (stablecoin market cap as a share of total crypto market cap), and on-chain velocity (transfer volume divided by supply). It also covers how stablecoins are spread across chains such as Ethereum, Tron and Solana.
The file gives threshold tables and simple rules, for example that a 30-day supply change above +5% is labelled strong bullish, and it combines the five indicators into a composite score from -10 to +10. The agent applies these rules to the data you provide.
Install the skill from the ZIP download
Download the ZIP file for stablecoin-flow from this page and unzip it. The skill is a folder that contains a file named SKILL.md. Place that folder in one of two locations:
For your personal use across projects, put it in ~/.claude/skills/. For a single project, put it in that project's .claude/skills/ folder.
Alternatively, you can upload the skill in the Claude apps under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/stablecoin-flow/
- Project: .claude/skills/stablecoin-flow/ inside your project
- Claude apps: Settings → Capabilities → Skills, then upload
Trigger the skill with natural requests
You do not need a special command. Ask in plain language about stablecoin supply, flows, reserves or liquidity. The skill's description covers stablecoin supply and flow analysis, USDT/USDC mint-burn signals, exchange stablecoin reserves, on-chain stablecoin velocity and capital rotation indicators for crypto market timing, so requests that use these ideas are the ones that should bring it into play.
Be specific about the data you are giving the agent. The skill does not fetch data, so a request works best when it includes the figures.
- Is there net USDT minting this week that could signal new capital entering crypto?
- Score the current stablecoin liquidity using these supply and exchange reserve numbers: [paste figures]
- What does a falling stablecoin dominance ratio usually indicate, according to the stablecoin framework?
- Compare USDC and USDT supply changes over 30 days and tell me what the USDC/USDT ratio suggests.
What you need before you start
The examples in the file are Python-style code snippets, so Python is the natural language for any calculations you do alongside the agent. The skill does not name a data source. You must supply current stablecoin supply, mint and burn, exchange reserve and on-chain transfer data yourself, or through another tool. The file lists possible sources such as DeFi Llama, Tether Transparency, Circle USDC Stats and Dune Analytics, and mentions a read_url tool for web-based sources, but it does not guarantee access to any of them.
- Python for running the calculations shown in the file
- Your own current data for supply, mint/burn, exchange reserves and transfers
- Optionally, access to a web-reading tool if you want the agent to pull from the listed sources
Safety notes: scripts and network
This skill runs no scripts and uses no network access, according to its listing. It is a plain instruction file (SKILL.md). That means it does not download prices or supply figures by itself. Any live numbers in the file, such as USDT at about $140B, are snapshots and will be out of date.
The signals are heuristic and the thresholds are unvalidated rules of thumb. The file presents correlations, such as BTC price and total stablecoin supply having r > 0.85, without sources. Do not treat the output as a trading or investment decision. The skill also does not cover regulatory, tax or accounting questions about stablecoins.
- Runs scripts: no
- Uses network: no
- Hardcoded figures are examples, not current data
- Not financial advice and not validated
Tips for useful results
Give the agent a date range and the source of every number, so the analysis is traceable. Ask it to show which threshold each indicator crossed, since the file's rules are explicit and easy to check. Ask for the composite score breakdown, so you can see which of the five components drove it.
If you want a repeatable report, ask the agent to follow the output format in the file, which uses a supply overview table with 7-day and 30-day changes.
- Always supply fresh numbers and a date range
- Ask which thresholds were triggered
- Request the component-by-component composite score
- Treat the output as context, not a signal to trade
What people use it for
- Writing a weekly market context note that tracks USDT and USDC supply changes against the file's threshold table.
- Scoring exchange stablecoin reserve changes over 7 days to label them as accumulation or deployment.
- Teaching students how stablecoin supply is used as a liquidity proxy and where the framework's assumptions are weak.
- Building a dashboard component that computes stablecoin dominance and labels it using the file's cut-offs.
- Comparing stablecoin distribution across chains such as Ethereum, Tron and Solana to describe where activity is moving.
Frequently asked questions
Is the stablecoin-flow skill free?
The skill is published under the MIT licence, as listed on this page. The data sources the file mentions have different access terms: some are free, while Glassnode, CryptoQuant and Nansen are listed as paid, with limited free tiers for some.
Do I need to install Python to use it?
The skill's examples are Python-style code, so Python is needed if you want to run those calculations. The skill itself is an instruction file and the listing says it runs no scripts.
Does the skill fetch live stablecoin prices or supply?
No. The skill does not fetch live data and uses network: no. The figures inside the file are example snapshots, so you must provide current data yourself.
Where do I put the skill folder?
Place it in ~/.claude/skills/ for personal use or in .claude/skills/ inside a project. You can also upload it in the Claude apps under Settings → Capabilities → Skills.
Can I use this skill to decide when to buy or sell crypto?
The file's signals are heuristic, its thresholds are unvalidated rules of thumb, and its correlations are presented without sources. It should not be used alone for trading or investment decisions.
How is this different from a stablecoin data dashboard?
A dashboard shows data. This skill gives the agent an interpretation framework with thresholds and a composite score. It does not replace a data source, so it works best alongside one.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you want an AI agent to interpret stablecoin supply and flows with a clear, written framework, and you are comfortable supplying your own current data. Pick a different tool if you need live figures, validated trading signals, or guidance on regulation, tax or accounting.
Download the skill