About tradeledger
tradeledger is a read-only view of the iBitLabs live trading experiment, for agents that need to report its actual state. It exposes account metrics, closed trades, rule status, and the experiment's narrative, without trading or account-write operations.
Key features
- 01Read balance, PnL, and win rate
- 02View recent closed trades
- 03List contributor rules by status
- 04Check a rule's hit rates
- 05Read the latest saga chapter
Use cases
- 01Verify the experiment's balance and open position
- 02Review recent trades and exit reasons
- 03Check rule status and promotion progress
- 04Read the experiment's latest narrative
MCP support
It forwards tool calls to a single upstream MCP endpoint and returns upstream results unchanged. The server exposes read-only data tools, with no order or account-write operation.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
5Saved tool list · updated 1w ago
This tool takes no parameters.
Details
From the project READMEtradeledger
A read-only window onto a live trading experiment. tradeledger lets an agent quote the experiment's actual state — its current balance and open position, the trades it has actually closed, the contributor shadow rules it runs, and the story it tells about itself — instead of inventing plausible numbers.
Nothing here trades. The five tools only read public data; there is no order, wallet, or account-write operation in the surface, and no credential is needed or accepted.
Tools
| Tool | Arguments | What it answers |
|---|---|---|
get_live_status | — | Balance, total PnL, win rate, trade count, market regime, and the open position if one is running. |
get_recent_trades | limit (optional integer) | The most recent closed trades, each with direction, exit reason, PnL, and regime. |
list_adopted_rules | status (optional: adopted, queued, or all) | The contributor-proposed shadow rules, filtered by where they stand. |
get_rule_status | rule_id (required string) | One rule's per-bucket hit rates and how far it sits from its promotion bar. |
get_latest_saga_chapter | lang (optional: en or zh) | The newest entry in the experiment's narrated saga. |
How it works
tradeledger is a forwarding client. A tool call is packaged as a
JSON-RPC tools/call request and sent to a single upstream MCP
endpoint; whatever the upstream returns is handed back untouched.
When the upstream fails, the answer is an error text beginning
tradeledger error: — never a locally invented figure.
Provenance
The endpoint and the forwarding shape are labelled reconstructions. The original server's public description names its data sources in prose but publishes no per-tool endpoint, so this rebuild points every call at one configurable upstream:
IBITLABS_MCP_URL— upstream MCP endpoint. Default:https://www.ibitlabs.com/api/mcp(labelled reconstruction).REQUEST_TIMEOUT_MS— per-request timeout. Default:15000.
Run it as an MCP server
{
"mcpServers": {
"tradeledger": {
"command": "npx",
"args": ["-y", "@reneturcios/tradeledger", "mcp"]
}
}
}
Or in a container:
docker run -i ghcr.io/cloudcorprecords/tradeledger
Run it as a CLI
npx @reneturcios/tradeledger get_live_status '{}'
npx @reneturcios/tradeledger get_recent_trades '{"limit": 2}'
npx @reneturcios/tradeledger get_rule_status '{"rule_id": "F"}'
npx @reneturcios/tradeledger --version
Each tool subcommand takes exactly one JSON object string with the tool's arguments and prints the result text. Exit codes are 0 on success, 1 on an error result, and 2 on a usage error.
Development
npm install
npm run build # tsc -> dist/ (dist/cli.js is the bin)
npm test # tsx --test test/*.test.ts — offline; the upstream is stubbed
npm run dev # tsx src/cli.ts — run the CLI from source
The test suite never touches the network: it drives the core
with a stubbed JSON-RPC upstream whose payloads are the public
worked-example responses, and checks they pass through
untouched, that failures become tradeledger error: results,
and that the served tool list matches the contracted five tools
exactly.
License
MIT. See LICENSE.
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.