Skip to main content
The MCP gives your agent live access to Allium’s blockchain data. Skills and plugins teach it what to do with that access — your team’s actual analytical workflow, encoded so the agent doesn’t have to relearn it every session.

​
Skills and plugins, briefly

  • A skill is a packaged set of instructions — following the open Agent Skills specification — that an agent loads when a task matches it, on top of whatever tools it already has.
  • A plugin is an installable bundle of skills for Claude Code, registered through a marketplace so a single command installs the whole set.
Allium ships two plugins from the allium-labs/skills repo: allium-agent (direct data access via the allium CLI and partner infrastructure skills) and allium-analyst (the analyst methodology covered on this page). You can install either independently.

​
The allium-analyst plugin

allium-analyst sits on top of the MCP tools you already have configured above. Instead of writing SQL directly against a guessed table, your agent works through a defensible methodology: pick the right source, prove it has coverage, build the query, validate the result, and save it so someone else can rerun it.

Match the right source

Compares candidate tables, dashboards, and endpoints against the actual question before any SQL gets written.

Investigate with a defensible method

Checks Allium Terminal for existing coverage, builds from a small test to a production query, and validates the result.

Sector-specific pitfalls

Applies known double-counting, classification, and aggregation mistakes for RWA, stablecoins, bridges, DEX, and lending.

Reproducible results

Saves load-bearing queries to Explorer instead of leaving the evidence only in the chat transcript.

​
Installation

You need the Allium MCP already configured before installing this plugin — see the Quick Start if you haven’t set that up yet.
1

Add the Allium Labs marketplace

In Claude Code, run:
2

Install the plugin

3

Restart Claude Code

Restart so the new skills load.
4

Try it

Ask something that needs the workflow, for example:
Use the Allium skills to investigate how tokenized RWA treasury supply has grown over the past 6 months, and save the query so I can share it.
If your agent checks Terminal for existing coverage, picks a real table instead of guessing, and hands you back a saved Explorer query at the end, it’s wired up correctly.
If you’re not on Claude Code, or want to install specific skills without the plugin system, use the generic Agent Skills installer instead:
Add --skill <name> (e.g. --skill data-matching) to install a single skill rather than everything in the repo.

​
Skills reference

allium-analyst bundles two core workflow skills and five sector-methodology skills. The sector skills plug into the workflow at the source-selection step — your agent reaches for one automatically once the question touches that sector.
Matches a question to the right Allium table, metric, dashboard, or realtime endpoint before any SQL gets written. Defines a data contract — object, measure, grain, scope, and time semantics — then compares candidate sources against it and proves coverage with a bounded test query rather than trusting that a chain or table merely appearing in search means it has real data. Runs first, before allium-investigation or any sector skill.
Runs the full investigation on top of the source data-matching selected. Checks Allium Terminal for existing coverage before writing anything from scratch, builds the query from a small ephemeral test to a saved production query, validates the result (reconciliation, duplicate-count checks, cross-checking at least one figure a different way), and separates observed fact from interpretation from unknown. Finishes by saving load-bearing queries to Explorer, with an optional dashboard hand-off prompt for Allium Assistant when the investigation produced multiple queries.
Sector playbook for tokenized real-world assets — treasuries, money market funds, tokenized equities, commodities, and private credit. Covers treating AUM/supply as a snapshot rather than summing it across days, correctly aggregating multi-chain fund deployments while disclosing the aggregation, classifying by underlying risk factor instead of legal wrapper, and excluding known bad-data tokens before ranking holders.
Sector playbook for stablecoin transfer volume, supply, holders, and geographic usage across USDC, USDT, DAI, and other issuers. Covers the raw-vs-adjusted volume distinction (raw can overstate real activity by 5-20x), chain-specific double-counting mechanics like XRPL’s trust-line rippling and Aptos’s dual token addresses, and address case-sensitivity rules that vary by chain.
Sector playbook for cross-chain bridge and messaging-protocol data — CCTP, LayerZero, Wormhole, Across, and canonical/application bridges. Covers the asymmetric, single-directional way transfers get recorded, the double-counting trap that shows up when filtering to two or more target chains at once, and how to trace a lock-and-mint transfer across chains when there’s no direct join key linking the two legs.
Sector playbook for DEX/AMM trades, liquidity, and pricing — Uniswap, PancakeSwap, Pump.fun, Orca, Pendle, aggregators, and perp DEXs like Hyperliquid. Covers per-hop double-counting on multi-hop routes, when to group by protocol versus project, and reconstructing an aggregator’s implicit fee when it shows up as a plain transfer instead of a labeled column.
Sector playbook for DeFi lending protocols — Aave, Morpho, Fluid, Euler, Maple, and forks. Covers joining the separate deposit/withdrawal/loan/repayment/liquidation event tables to detect operations with no dedicated label (collateral swaps, deleveraging), column-naming inconsistencies across those tables, and fill-level double-counting on liquidations.

MCP Overview

Set up the Allium MCP server itself

Tools Reference

Detailed documentation for the MCP tools these skills call

allium-labs/skills

Skill source, README, and the allium-agent plugin

Examples

Real interaction transcripts