About TradeOS Public Intelligence
TradeOS Public Intelligence is a read-mostly MCP server for agents and MCP clients to access TradeOS public crypto-market research. It provides market analysis, evidence, watchlist data, and feedback tools without trading or exposing private telemetry.
Key features
- 01Market digests and public candidates
- 02Thesis and token watchlists
- 03Evidence packets and claim proofs
- 04Feedback with provenance fields
- 05Read-only access; no trade execution
Use cases
- 01Review public crypto-market research
- 02Inspect evidence behind a market claim
- 03Track saved watchlists with an account token
MCP support
Exposes TradeOS public-intelligence tools to stdio MCP clients, including Claude Desktop and Cursor. Some watchlist tools require an account token.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @agenticsrclab/tradeos-public-intel-mcp-serverTRADEOS_API_BASEOptional TradeOS public intelligence API base URL.TRADEOS_PUBLIC_INTEL_KEYSecretOptional public-intel app key for attribution and quota confidence.TRADEOS_ACCOUNT_TOKENSecretOptional account token for user-owned watchlist tools.TRADEOS_X402_BASE_URLOptional TradeOS x402 gateway base URL.TRADEOS_X402_AGENT_IDOptional stable agent or host identifier used for x402 attribution.TRADEOS_X402_PAYMENT_HEADERSecretOptional one-time buyer-signed x402 payment header for explicit paid calls. Do not use for private keys.TRADEOS_X402_WALLET_ADDRESSOptional buyer wallet address sent with explicit paid x402 calls.Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
29Saved tool list · updated 2w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| limit | integer | No | |
| chainId | string | No | |
| windowEnd | string | No | |
| windowStart | string | No |
Details
From the project README@agenticsrclab/tradeos-public-intel-mcp-server
Read-mostly MCP server for the TradeOS public Data Intelligence layer.
Official MCP Registry name:
io.github.agenticsrclab/tradeos-public-intel-mcp
TRADEOS_API_BASE=https://api.tradeos.tech/v1/public-intel \
TRADEOS_ACCOUNT_TOKEN=<optional-account-token-for-watchlist-tools> \
TRADEOS_PUBLIC_INTEL_KEY=<optional-app-key-for-attribution> \
npx @agenticsrclab/tradeos-public-intel-mcp-server
Use it from Claude Desktop, Cursor, or another stdio MCP client to give an agent access to:
tradeos.get_market_digest
tradeos.get_public_candidates
tradeos.get_thesis_watchlist
tradeos.get_symbol_thesis
tradeos.get_evidence_packet
tradeos.get_public_claim_proof
tradeos.watchlist_capabilities
tradeos.get_token_watchlist_snapshot
tradeos.list_watchlists
tradeos.create_watchlist
tradeos.add_watchlist_item
tradeos.get_watchlist_state
tradeos.list_watchlist_events
tradeos.list_watchlist_deliveries
tradeos.trigger_watchlist_deliveries
tradeos.submit_digest_feedback
tradeos.submit_thesis_feedback
tradeos.submit_claim_outcome_feedback
tradeos.submit_watchlist_feedback
Feedback tools accept optional provenance fields:
feedbackSource: human | human_assisted | agent | automation | hybrid
automationLevel: none | assisted | automated | autonomous
agentId
agentRunId
agentModel
agentConfidence
tradeos.get_credit_state can return app reputation DTI for the configured
public-intel app key. It does not expose personal human DTI; human DTI remains a
signed-in TradeOS dashboard and Review Lab lifecycle.
The server does not place trades, accept exchange credentials, or expose private TradeOS telemetry.
Access model:
Free public kit: bounded reads, token snapshots, and feedback writes
Human DTI: public dashboard depth, public Ask packs, or read-only Review Lab where enabled
App reputation DTI: app-key feedback quality and quota confidence, not personal balance
Account token: saved watchlists, events, channels, delivery audit, and user-owned feedback
Paid TradeOS/x402: automation, exports, high-volume alerts, premium data, validation APIs
TRADEOS_PUBLIC_INTEL_KEY is optional and used only when TradeOS has issued a
public-intel app key for attribution. This MCP server can use the key, but it
does not create keys.
Watchlist tools require TRADEOS_ACCOUNT_TOKEN. They expose research awareness,
saved user state, and delivery audit only; they do not place trades or connect
exchanges.
Learn more:
Homepage: https://tradeos.tech
Public docs: https://tradeos.tech/llms.txt
Key setup: https://github.com/agenticsrclab/tradeos-public-intel-kit/blob/main/docs/getting-api-keys-and-scale.md
x402 discovery: https://tradeos.tech/.well-known/x402.json
Official MCP Registry Publishing
The registry metadata is in server.json. The package mcpName in
package.json must stay equal to server.json.name.
Local metadata check:
npm --workspace @agenticsrclab/tradeos-public-intel-mcp-server run registry:check
Publish prerequisites:
GitHub namespace: io.github.agenticsrclab/*
NPM package: @agenticsrclab/tradeos-public-intel-mcp-server
MCP registry name: io.github.agenticsrclab/tradeos-public-intel-mcp
Manual publish from this directory after the npm package version is public:
mcp-publisher validate
mcp-publisher login github
mcp-publisher publish
CI publish uses .github/workflows/publish-mcp-registry.yml and GitHub OIDC for
the MCP Registry. The workflow still needs an NPM_TOKEN repository secret when
it is also publishing the npm package version. The npm package uses the
@agenticsrclab scope because that scope is controlled by the same public
distribution org used for GitHub registry authentication.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.