Skip to main content
Glama

start_paper_trade

Destructive

Start a strategy live: it runs on live market data, on paper, with a note on it every morning (the lab's paper trade). It is what "start it live, follow it, paper trade it, run it live, watch it" ask for, and what a tap on a run's offer to start it live does. save_trading_config never does this: it keeps a run's settings, and nothing goes live from it.

    A first call starts nothing: its answer carries `ask`, the question for the person, with two offers. The same
    call with confirm=true, made after the person's yes, starts it.

    What it tracks: a finished run of this account (task_id: a ready test, a paper's circuit, a saved lab
    project's run, a backtest or a walk-forward), a published paper (study, from search or fetch) or one of the
    person's saved lab projects (project_id, from list_my_primitives_projects).

    It registers the strategy in the person's saved work, sets it up on paper with 100,000 for 90 days, checks it
    on the last six months and puts it live. Its first update comes after the next US market close. When the
    check takes longer, it goes on on QuanterLab and the strategy goes live by itself when it is done. Nothing
    here places an order or moves money: it is simulated on paper, on the data served each day.

    An account runs one strategy at a time. With one already tracked, a call starts nothing and its `ask`
    offers the swap, with two offers. The same call with replace=true and the replace_id its message names,
    made after the person's yes, stops that one (its record is kept) and tracks this one instead.

    Its card shows what started; `message` and `open_url` say the same for a host without cards. A strategy that
    cannot go live as it is answers with why in `message`, and, for a paper whose data stops at its publication,
    an offer with the exact call of its live version. my_paper_trades says later how it is doing.

    Args:
        task_id: The task_id of one of this account's finished runs.
        study: Instead of task_id, a published paper's id.
        project_id: Instead of task_id, the id of one of the account's saved lab projects.
        live_version: With study, true tracks the paper's live version: the same rule on the data served
            today.
        replace: True once the person has said yes to tracking this one instead of the one the account
            tracks.
        replace_id: With replace, the id of the strategy the account tracks, as the answer that asked names it.
            Another one tracked by then is asked about again.
        confirm: True once the person has said yes to the question the first call asked.
    

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "start_paper_trade"
  }
}

Input Schema

NameRequiredDescriptionDefault
studyNo
confirmNo
replaceNo
task_idNo
project_idNo
replace_idNo
live_versionNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.9/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations declare destructive/openWorld/non-idempotent, and the description adds substantive behavior on top: the first call deliberately starts nothing, an account runs one strategy at a time, replace stops the existing strategy but keeps its record, setup uses 100,000 for 90 days with a six-month check, the first update lands after the next US market close, a slow check defers to QuanterLab and auto-goes-live, and no orders or money move. It also explains failure surfacing through `message` and `open_url`. This is exactly the mutation/async context annotations cannot carry.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the purpose and the save_trading_config exclusion before protocol details, which is the right order. It is long and occasionally repetitive (the user-phrase alias list, the restatement of the confirm flow in both prose and Args), but nearly every paragraph carries distinct routing or behavioral information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 7-parameter, destructive, open-world, asynchronous mutation with no output schema, the description covers the full contract: two-phase confirmation, replacement conflict handling, async deferral and auto-go-live, response shape (`ask`, `message`, `open_url`, card), and the failure path with an offered live-version call. Nothing an agent needs in order to invoke it correctly is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0% (bare generic titles, no descriptions), so the description must carry all seven parameters — and it does, through the Args block. It adds semantics the schema cannot express: task_id/study/project_id are mutually exclusive sources ('Instead of task_id'), live_version only pairs with study, replace_id must come from the prior answer and goes stale ('Another one tracked by then is asked about again'), and confirm is tied to the preceding ask.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Start a strategy live... it runs on live market data, on paper') and immediately distinguishes itself from the nearest sibling by naming save_trading_config and declaring what it never does. An agent can tell this apart from start_test, start_backtest, or save_trading_config without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit when-to-use context ('start it live, follow it, paper trade it'), explicit when-not (save_trading_config keeps settings only), and the exact two-step protocol: first call returns `ask`, second call with confirm=true after the person's yes. It also names the alternative for later status checks (my_paper_trades) and the swap path with replace/replace_id.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources