Learn how to operate Miwa like a pro

1. Access the dashboard: go to the hosted URL (or your self-hosted instance), create an account with your invite code, and connect your wallet. Keep at least 0.05 SOL available for buys and fees.
2. Create a strategy preset: under Strategies → New Strategy choose a template, set entry filters (bonding-curve range, minimum holders, liquidity), configure buy size/slippage/priority fee, and add exit rules for take profit + stop loss.
3. Spin up the bot: Bots → New Bot, pick the strategy and wallet, set budget caps and optional modules like Discord alerts, then click Start.
4. Monitor: watch the Activity Feed or Discord webhook for detections, buys, and sells. Pause or clone the bot any time.

After the first execution cycle, open the Logs tab to review RPC latency, transactions per second, and any warnings (e.g., slippage or liquidity rejections). Export the CSV if you need to provide proof of trades to teammates and annotate which presets are safe to scale.

• Invite-only access is enforced—store your promo code securely since it is linked to billing and support.
• Keep a hardware wallet in read-only mode for manual overrides; you can hot-swap wallets without stopping the bot by pausing and reassigning from Settings → Wallets.
• Schedule a weekly cadence to rotate RPC keys and rotate Discord webhooks to minimize leaked credentials.

Each preset is split into three blocks: Entry filters (who qualifies), Execution (how much SOL, slippage, fee profile), and Exit (profit/loss targets, max hold time, safety triggers). Build multiple presets for different market conditions.

• Filters: e.g., bonding curve 5–35%, minimum 15 holders, ≥2 SOL real liquidity, blacklist risky deployers.
• Execution: 0.002–0.01 SOL per buy, 0.35–0.45 slippage, “Balanced” (150k µLamports) priority fees by default.
• Exit & risk: TP 4×, SL 20%, auto-sell if liquidity collapses, schedule trading windows, cap concurrent positions.

Clone presets before big edits, label them (“production”, “experimental”), and rely on the built-in audit log to track changes across operators.

To avoid overlapping trades, create separate presets for each wallet and cap concurrent positions to your risk appetite. Operators commonly run “alpha”, “momentum”, and “manual confirm” presets side by side— staggering them ensures liquidity is spread across tokens while still catching new launches.

Use the dry-run mode when testing a new preset: Miwa simulates entries, shows predicted fills, and logs the filters that rejected each candidate. You can then tighten or loosen parameters without burning SOL.

1. Dry run: duplicate the preset, shrink buy amount to 0.001 SOL, and run the bot in observation mode to ensure detections align.
2. Go live: enable the bot with the small buy size, confirm buy + sell transactions on Solscan, and ensure Discord/web alerts fire.
3. Validate guardrails: max concurrent tokens, auto-rug detection, and schedules should all trigger as configured.
4. Scale gradually: only increase size after a few clean trades; add more bots/presets once confident.

Capture screenshots (or screen recordings) of the initial trades so anyone reviewing later can see the parameters you used, the RPC endpoint health, and proof that safeguards fired. This artifact is useful for internal audits or when asking Miwa support to replicate what you saw on-chain.

The telemetry drawer highlights latency spikes and RPC retries; unusual spikes usually mean you should rotate to a backup RPC or temporarily reduce buy pressure until the network stabilizes.

Miwa ships with presets such as Moonshot Hunter (aggressive early entries), Scalper (tight TP/SL), and Manual Assist (you confirm before execution). Each module consists of reusable blocks—filters, execution plans, and exit trees—that you can rearrange.

• Filters: holder velocity, bonding curve %, liquidity floor, blacklist tags.
• Execution: auto-priority fees, anti-sandwich padding, delayed entries.
• Automation extras: circuit breakers, daily profit targets, schedule windows.

Combine these to produce custom behaviors (e.g., a conservative weekend preset vs. aggressive launch-hours preset).

Teams often dedicate specific wallets to a strategy archetype: one wallet constantly scalps and automatically compounds profits while another wallet only fires on curated allowlists. Use Miwa’s coordination module to enforce these boundaries so capital does not overlap.

Every block is versioned, so when Miwa ships an update you can diff what changed (e.g., new rug heuristics) and roll back if your custom logic behaves differently.

• Bot won’t start: ensure wallet has SOL, access code is valid, and no other bot uses the same wallet.
• No detections: relax filters or confirm Pump.fun is online and inside your schedule window.
• Orders failing: increase slippage/priority fees or lower buy amount if funds are insufficient.

The Activity → Diagnostics tab mirrors these states with time stamps; look for red “blocked” badges to see whether rate limits, wallet caps, or paused schedules are the culprit. Copy the diagnostic hash when contacting support—they can replay the exact run in staging.

Codes follow the MODULE-CATEGORY format. Filter the run history in Miwa via “Copy error code” to view every occurrence along with the strategy + wallet combo that triggered it.

1. Create a webhook in Discord (Server Settings → Integrations) and copy the URL.
2. In Miwa, go to Settings → Notifications, paste the webhook, and send a test event.
3. Toggle which events you want (bot lifecycle, detections, orders, risk alerts).

If alerts stop, regenerate the webhook and ensure the Discord channel allows external posts.

Power users connect multiple webhooks: one for high-signal detections, another for accounting exports, and a private one for pager-duty style warnings. Prefix alerts with emojis (“🚀 Buy executed”) so it is easy to glance which type fired.

• Is Miwa custodial? No, you sign with Phantom locally.
• How many bots? As many as you want—just use a unique wallet per bot.
• Pump.fun downtime? Bots pause automatically and notify you on Discord.
• Can I self-host? Yes, docker-compose templates are in dashboard/; bring your own RPC keys and run behind Cloudflare or Tailscale.
• Do you support Solflare? Phantom is the default; a Solflare connector is in beta—reach out on Discord if you want to join the test cohort.

The full policy and SLA live inside docs/ of the repo. When in doubt, open a ticket through the dashboard—responses average under half a day.