ask_turbine
ChatGPTAsk Turbine's AI to explain the authenticated account, strategy, backtest, deployment, or trading evidence. It may propose actions but cannot confirm or execute them.
cancel_research
ChatGPTCancel an owned running Deep Research job. Apply only after the user explicitly approves the exact preview from preview_cancel_research. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
copy_session
ChatGPTCreate a separate saved copy of one owned Studio session, preserving the original. Apply only after the user explicitly approves the exact preview from preview_copy_session. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
copy_shared_strategy
ChatGPTCopy one published library strategy into a new owned Studio draft. Apply only after the user explicitly approves the exact preview from preview_copy_shared_strategy. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
create_research
ChatGPTSave a new owned Kalshi crypto or weather research draft using the exact thesis, base DSL, and optional parameter grid stored in its approved preview. Returns the saved report details; use its ID as report_id to read the report or request a separate research run. Does not start research, execute trades, or deploy a bot. Apply only after the user explicitly approves the exact preview from preview_create_research. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
create_strategy
ChatGPTGenerate and save a new owned strategy draft from the platform and message stored in its approved preview. May consult public market information while generating the draft. Returns the saved session_id and available draft details for subsequent inspection. Does not deploy or activate a bot, submit orders, or move funds. Apply only after the user explicitly approves the exact preview from preview_create_strategy. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
delete_strategy
ChatGPTDelete one owned saved strategy and its Studio session. This removes saved draft data. Apply only after the user explicitly approves the exact preview from preview_delete_strategy. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
get_account_status
ChatGPTCheck the authenticated account's membership, runner setup, and usage limits.
get_analytics
ChatGPTRead the performance summary for the authenticated user's bot_id, including available profit/loss, trade counts, and activity metrics. Use it to explain observed bot performance. Coverage depends on available runner and journal evidence; missing data is not proof of zero activity. Does not place orders or control the bot.
get_backtest
ChatGPTRead the saved historical backtest result for an owned session_id from list_sessions or list_strategies. Returns the stored result, strategy DSL hash, platform, and simulation caveats when available. Use it to explain simulated outcomes; a session without a saved result reports not found. Does not start or rerun a backtest.
get_backtest_job
ChatGPTRead the status of an owned historical backtest job using its backtest_id returned when the job was started. Returns the job status and available result references or failure information. Use it to check whether a requested simulation is still pending or finished; pending status is not a completed result. Does not start or retry a job.
get_deployment_state
ChatGPTRead deployment state for bot_id from a Studio strategy. A strategy with no deployment may have no state.
get_execution_diagnostics
ChatGPTInspect execution diagnostics for an owned bot_id, combining saved run state with a runner probe when available. Use it to investigate current status, setup problems, and differences between saved deployment state and observed runner activity. Does not restart, deploy, stop, or change a bot.
get_journal_contracts
ChatGPTRead the authenticated user's recorded contracts for run_id, including per-contract activity and results. Supports cursor, limit, from, and to filters. This does not place orders or modify the run.
get_journal_events
ChatGPTRead recorded execution events for an owned run_id from list_journal_runs. Returns an events page and next_cursor when more records exist; use cursor and limit to continue, order to choose direction, and from/to to restrict the recorded time window. Use these observations to explain a run's activity. Does not place orders or modify the run.
get_journal_run
ChatGPTRead one recorded run by its owned run_id from list_journal_runs. Returns the strategy snapshot, run status, available performance and contract summaries, and coverage information. Use it to establish what was recorded before interpreting events or results; incomplete coverage is not evidence of no trading. Does not start or control a run.
get_journal_samples
ChatGPTRead recorded equity and exposure samples for the authenticated user's run_id. Supports limit, from, and to filters. Returns historical observations without modifying the run or executing trades.
get_library_strategy
ChatGPTRead a public library strategy by strategy_id (ID or slug) from list_strategy_library, including DSL rules, parameters, performance, and available backtest details. Use this to explain a strategy shape and compare variants.
get_paper_state
ChatGPTRead the latest recorded paper session for an owned bot_id, including its active flag, session statistics, and daily usage when available. A bot without a recorded paper session returns no session. Use it to inspect existing simulated activity. Does not start, stop, or activate paper trading.
get_pnl_history
ChatGPTRead the recorded profit and loss history for an owned bot_id. Returns available historical P&L points and a coverage caveat for interpreting the series. Use it to explain changes over the recorded period; an empty history means no recorded points, not a verified zero return. Does not place orders or move funds.
get_public_research_report
ChatGPTRead a public Deep Research report by slug, including its thesis, analysis, parameter comparisons, and strategy variants.
get_research_report
ChatGPTRead one owned research report by report_id from list_research_reports.
get_runtime_logs
ChatGPTRead available deployment/runtime logs scoped to the authenticated user's bot_id. Use them to investigate reported runtime errors or missing activity. Missing logs do not establish that a bot is healthy or inactive. Does not restart a runner, deploy code, or change bot state.
get_session
ChatGPTRead one owned Studio session using session_id from list_sessions or list_strategies. Returns its summary, saved messages, configuration snapshot, strategy DSL when present, and model information. Use it to inspect the saved draft and its conversation. Does not modify the session or execute the strategy.
get_strategy
ChatGPTRead one owned Studio strategy by strategy_id from list_strategies. Includes the saved DSL and session state.
get_trades
ChatGPTRead recorded trades for an owned bot_id, with limit controlling the number of returned records. Returns available trade records and a coverage caveat. Use it to inspect observed executions; these are records of existing activity, not order instructions. Does not place or cancel orders.
list_journal_runs
ChatGPTList the authenticated user's recorded trading journal runs. Filter by mode and status; use limit and cursor for pagination. Each summary supplies run_id for get_journal_run and the event, contract, and sample read tools. Use it to locate recorded live or paper activity. Does not start or control any run.
list_paper_equity
ChatGPTRead recorded equity history for the latest paper session of an owned bot_id, with limit bounding returned snapshots. Returns the paper session ID, starting balance, and available history; no recorded session produces an empty history. Use it to explain simulated account value over time. Does not start or alter paper trading.
list_paper_fills
ChatGPTRead recorded simulated fills for the latest paper session of an owned bot_id, using limit to bound returned records. Returns the paper session ID when available and its recorded fills, or an empty list if no session exists. Use it to inspect historical paper executions. Does not place orders or start paper trading.
list_public_research_reports
ChatGPTBrowse completed public Deep Research reports from /studio/research, newest completed first, with cursor pagination. Pass a returned slug to get_public_research_report.
list_research_reports
ChatGPTList compact summaries of recent owned research reports. Use get_research_report for complete results.
list_run_history
ChatGPTRead the saved execution run history for an owned bot_id. Returns recorded runs and their available lifecycle details; a strategy without recorded runs returns an empty list. Use it to distinguish past runs from current execution diagnostics. Does not start, stop, or redeploy a run.
list_sessions
ChatGPTList owned Studio sessions with pagination. Current Studio strategy IDs also identify their sessions.
list_strategies
ChatGPTList or search owned Studio strategies. Pass query to search names; use an item id as strategy_id and, for current Studio deployments, as bot_id.
list_strategy_library
ChatGPTBrowse the public strategy library at /studio/strategies. Search with query; sort by recent (newest), name, pnl, roi, sharpe, win_rate, stars, or market. Set group=family to discover strategy shapes, then pass family to compare their variants. Supports venue, category, asset, interval, engine, and market filters plus cursor pagination. Library IDs are for get_library_strategy or copy_shared_strategy; they are not deployed bot IDs.
preview_cancel_research
ChatGPTValidate and preview: Cancel an owned running Deep Research job. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling cancel_research.
preview_copy_session
ChatGPTValidate and preview: Create a separate saved copy of one owned Studio session, preserving the original. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling copy_session.
preview_copy_shared_strategy
ChatGPTValidate and preview: Copy one published library strategy into a new owned Studio draft. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling copy_shared_strategy.
preview_create_research
ChatGPTPreview saving a research draft from market_type (crypto or weather), thesis, and base_dsl, with optional Kalshi venue and numeric sweep_dims. Returns a summary, warnings, resource versions, expiration, and confirmation_token for create_research. Does not save the research draft or start research; only a temporary confirmation receipt is stored.
preview_create_strategy
ChatGPTPreview creating a strategy draft from the required platform and task-specific message, with an optional arbitrage_leg_b_platform. Returns a summary, warnings, resource versions, expiration, and confirmation_token for create_strategy. Does not save or generate the strategy, deploy a bot, or place orders; only a temporary confirmation receipt is stored.
preview_delete_strategy
ChatGPTValidate and preview: Delete one owned saved strategy and its Studio session. This removes saved draft data. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling delete_strategy.
preview_start_backtest
ChatGPTValidate and preview: Start a historical simulation for one owned strategy, using the requested data window and optional simulated balance and limits. Stores simulation results; simulated trades are never submitted orders. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling start_backtest.
preview_start_research
ChatGPTValidate and preview: Start a stateful Deep Research job for one owned report. Runs historical parameter experiments and analysis that may consult public sources, consumes applicable account usage, and saves results. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling start_research.
preview_update_strategy
ChatGPTValidate and preview: Revise one owned saved strategy using a task-specific instruction. Overwrites draft rules and may consult public market information. Does not change an active runner. Stores a short-lived confirmation receipt without applying the operation. Present the exact parameters, summary, and warnings for approval before calling update_strategy.
rename_session
ChatGPTRename one owned Turbine Studio session. Overwrites its saved display name. This idempotent change does not require preview.
start_backtest
ChatGPTStart a historical simulation for one owned strategy, using the requested data window and optional simulated balance and limits. Stores simulation results; simulated trades are never submitted orders. Apply only after the user explicitly approves the exact preview from preview_start_backtest. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
start_research
ChatGPTStart a stateful Deep Research job for one owned report. Runs historical parameter experiments and analysis that may consult public sources, consumes applicable account usage, and saves results. Apply only after the user explicitly approves the exact preview from preview_start_research. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.
update_strategy
ChatGPTRevise one owned saved strategy using a task-specific instruction. Overwrites draft rules and may consult public market information. Does not change an active runner. Apply only after the user explicitly approves the exact preview from preview_update_strategy. Accepts only that operation's one-time confirmation token; uses its stored parameters. This tool cannot place orders, deploy bots, activate runners, or move funds.