Skip to main content
πŸ“ˆ Intermediate

Bot Lifecycle Management Best Practices

How to manage drafts, active bots, and archived strategies in RelayDesk as your portfolio of automations grows.

Running one bot is straightforward. Running fifteen is a different problem entirely.

As you develop more strategies and iterate on configurations, your RelayDesk workspace can quickly accumulate drafts, experimental setups, seasonal strategies, and deprecated approaches β€” alongside your live, production automations. Without a clear system for managing these, it becomes hard to know what's running, what's paused, and what's just noise.

This guide covers RelayDesk's bot states, when to use each one, and the organizational practices that keep your workspace manageable as you scale.

Understanding Bot States

Every bot in RelayDesk exists in one of four states. Understanding what each state does β€” and doesn't do β€” is the foundation of good lifecycle management.

Draft

A Draft bot has been created and configured but has never been activated. Drafts do not listen for signals and cannot execute trades.

Use Draft when:

  • You're building a new strategy configuration and not ready to paper trade yet
  • You want to save a configuration template you might activate later
  • You're mid-way through setup and need to return to it

Draft bots appear in your bot list but are visually distinguished from active bots. They have no signal history and no trade log entries.

Active

An Active bot is live. It listens for incoming signals and will execute trades according to its configuration. Active bots can be set to Live (real money) or Paper (simulated) mode β€” but in both cases, they are consuming your signal routing and are fully operational.

Use Active when:

  • The strategy is validated and ready to run
  • You're paper trading to confirm live behavior before funding
  • The bot is in a monitored, production state
TGIP Principle: RelayDesk recommends running any new bot in Paper mode first β€” even strategies you've traded manually before. Automation surfaces edge cases that manual trading never encounters.

Paused

A Paused bot retains its full configuration and history but stops processing incoming signals. Any signals sent to a paused bot will appear in Signal History with a status of ignored.

Use Paused when:

  • You're in a planned no-trade period (earnings, macro events, low-liquidity windows)
  • A strategy needs review but you don't want to lose its configuration or history
  • You're temporarily running a competing strategy on the same instrument
  • The market regime has changed and you're not yet ready to adapt the strategy

Paused bots are easy to restart β€” toggle them back to Active with a single click.

Archived

An Archived bot is removed from your active workspace but preserved with its full history intact. Archived bots cannot receive signals and do not appear in your main bot list. They are accessible via Bots β†’ Archived.

Use Archived when:

  • A strategy has been retired but you want to retain its Signal History and Trade Log for analysis
  • A seasonal strategy (e.g., earnings plays, specific market conditions) should hibernate until conditions return
  • You're cleaning up your workspace without permanently deleting configurations

Archiving is non-destructive. An archived bot can be unarchived and reactivated at any time.

When to Delete vs. Archive

Deletion is permanent. Once a bot is deleted, its configuration, Signal History, and Trade Log are gone.

Delete a bot when:

  • It was created by mistake and has no meaningful history
  • It's a duplicate of another bot with identical configuration
  • The data it holds has been exported and you have no reason to reference it again

Archive a bot when:

  • It has trade history worth preserving
  • You might want to revisit the configuration
  • You're unsure β€” when in doubt, archive rather than delete

A good default rule: if a bot has ever been Active, archive it rather than delete it.

Naming Conventions That Scale

When you have 3 bots, naming conventions don't matter much. At 15+, inconsistent naming becomes a real operational problem. A consistent naming system prevents you from accidentally modifying the wrong bot and makes your workspace easier to understand at a glance.

A practical naming structure for RelayDesk bots:

[Instrument]-[Strategy]-[Direction]-[Mode]

Examples:
SPY-MomentumScalp-Long-Live
QQQ-MeanReversion-Both-Paper
AAPL-EarningsPlay-Short-Paper
ES-BreakoutSystem-Long-Live

Breaking this down:

  • Instrument β€” The ticker or underlying (SPY, QQQ, ES, NQ)
  • Strategy β€” A short descriptor of the approach (Scalp, Breakout, MeanRev)
  • Direction β€” Long, Short, or Both
  • Mode β€” Live or Paper

This naming pattern makes your bot list instantly readable and makes it obvious which bots are in production versus testing.

Managing Drafts

Drafts accumulate faster than any other bot state. It's easy to start configuring a new idea, get distracted, and end up with six incomplete drafts cluttering your workspace.

A few practices that keep drafts manageable:

Set a completion or deletion deadline. When you create a draft, give it a clear next step. Either complete it within the session or note why it's pending. Drafts that have sat untouched for 30+ days are usually safe to delete.

Use the bot description field. RelayDesk's bot description field isn't just decoration β€” it's a note to your future self. Record what you were building, what's incomplete, and why you paused configuration.

Batch-review drafts weekly. A quick weekly pass through your Draft list β€” activate, complete, or delete β€” prevents a backlog from building.

Running Multiple Active Bots on the Same Instrument

Running two or more active bots targeting the same instrument is intentionally allowed in RelayDesk. There are valid reasons to do this: running a long-biased bot and a short-biased bot on the same underlying simultaneously, or layering a scalp strategy alongside a swing strategy.

A few things to manage carefully when doing this:

Signal routing is per-bot, not per-instrument. Each bot has its own webhook endpoint. Make sure your TradingView alerts are routed to the correct bot β€” it's easy to accidentally send signals to the wrong bot when you have multiple configured for the same ticker.

Position sizing stacks. If both bots enter simultaneously on the same instrument, your combined exposure is the sum of both bots' sizing. Factor this into each bot’s position-sizing configuration.

Pause one before debugging the other. If unexpected behavior occurs and you’re not sure which bot is responsible, pause all but one bot and isolate the issue in Signal History before reactivating the others.

Archiving Seasonal or Regime-Specific Strategies

Some strategies are best suited to specific market conditions β€” high VIX environments, earnings seasons, and specific economic cycles. Rather than deleting these between relevant windows, archive them with a clear naming convention:

[Instrument]-[Strategy]-[Condition]-ARCHIVED

Examples:
SPY-0DTE-HighVIX-ARCHIVED
AAPL-EarningsCall-Q1-ARCHIVED

When conditions return, you can unarchive, review the configuration against current conditions, and reactivate. This preserves institutional memory in your automation setup β€” you don’t have to rebuild from scratch every time.

Exporting History Before Archiving

Before archiving any bot with significant trade history, export its logs. Strategy analysis compounds over time β€” having 6 months of Signal History and Trade Log data in CSV form lets you run performance attribution, identify regime-dependent behavior, and inform future configuration decisions.

To export: Bots β†’ [Select Bot] β†’ Signal History β†’ Export CSV and repeat for the Trade Log. Store these alongside the bot's strategy notes.

A Practical Workspace Audit Checklist

Run through this monthly to keep your workspace clean:

  • Review all Draft bots β€” complete, document, or delete each one
  • Review all Paused bots β€” confirm they're paused intentionally and have a plan to reactivate or archive
  • Confirm all Active bots are live intentionally β€” no forgotten paper tests running
  • Export trade history for any bot you plan to archive
  • Archive retired strategies rather than leaving them paused indefinitely
  • Verify naming conventions are consistent across all active bots

Congratulations! πŸŽ‰

You've completed this tutorial. Mark it as complete to track your progress.