Base Trader

Autonomous crypto trading on Base via Bankr. Use for trading tokens, monitoring launches, executing strategies, or managing a trading portfolio. Triggers on "trade", "buy", "sell", "launch", "snipe", "profit", "PnL", "portfolio balance", or any crypto trading task on Base.

Security Scan

The skill's behavior mostly matches a trading helper that delegates to an external Bankr script, but there are multiple coherence issues (undocumented credential handling, inconsistent paths, undeclared binary deps and unknown source) that deserve caution before installing or running autonomously.

Purpose & Capability

The skill's stated purpose (autonomous trading via Bankr) is plausible and its instructions call Bankr commands to execute trades. However the skill metadata declares no required credentials or env vars while the runtime instructions explicitly rely on a Bankr config and an on-chain wallet (e.g., ~/.clawdbot/skills/bankr/config.json and ~/clawd/skills/bankr/scripts/bankr.sh). Delegating credentials to another skill (Bankr) is reasonable, but the lack of any declared relationship in metadata and inconsistent paths ('.clawdbot' vs 'clawd') is an incoherence that obscures what credentials/folders it will access.

Instruction Scope

SKILL.md instructs the agent to run external Bankr scripts at hardcoded paths, to read and write local data files (data/trades.json, data/performance.json), and to perform autonomous periodic actions (cron/heartbeat flows). It does not instruct reading arbitrary system files, but it assumes access to another skill's scripts/config and to wallet credentials that are not documented in the metadata. Several paths are inconsistent across files (e.g., ~/.clawdbot vs ~/clawd), which could cause misexecution or indicate sloppy packaging. The instructions give the agent broad discretion to run Bankr commands (buy/sell/set stop loss), which means if Bankr config contains keys, the skill can trade on the user's behalf.

Install Mechanism

There is no install spec (instruction-only plus two tiny scripts). That minimizes installer risk because nothing is downloaded at install time. The included shell scripts are small and readable; no network download/install steps are present in the repo itself.

Credentials

Metadata lists no required environment variables or credentials, yet SKILL.md requires a Bankr config file and an on-chain wallet to be present. The skill relies implicitly on credentials stored by the Bankr skill (or in the referenced config path) but does not declare or surface this in requires.env/primaryEnv. Additionally, the log script uses jq but the skill does not declare jq or other binary dependencies. This mismatch reduces transparency about what secrets and system state the skill will access.

Persistence & Privilege

The skill does not request always:true and does not modify other skills' configs. It runs autonomously if invoked (disable-model-invocation is false), which is the platform default — combine that with the ability to call trading commands and you'll get automated trading behavior, but the skill itself does not demand elevated platform persistence.

What to consider before installing

What to consider before installing/using this skill: - Transparency of credentials: The skill expects a Bankr configuration and an on-chain wallet but declares no required credentials. Inspect the Bankr skill and the config file (~/.clawdbot/skills/bankr/config.json or ~/clawd/skills/...) to see what secrets (API keys/private keys) it stores. Do not point it at a config that contains unlocked private keys unless you trust the Bankr skill and the environment. - Path inconsistencies: SKILL.md and README reference multiple different paths ('.clawdbot' vs 'clawd' and both ~$HOME and ~/clawd). Verify and correct these paths before running; mismatches could cause accidental use of different files or failure. - Autonomy risk: The skill instructs the agent to run bankr.sh to execute buys/sells. If you allow autonomous invocation, the agent could place real trades using whatever credentials Bankr exposes. If you keep it, prefer manual invocation or limit its permissions initially. - Inspect Bankr and bankr.sh: This skill delegates all network/trading actions to an external script (bankr.sh). Confirm that the bankr.sh you have installed is the legitimate Bankr integration you expect; inspect that script for endpoints and credential usage. - Small runtime tests: Before trusting live trading, run the check-portfolio and log-trade scripts manually to verify behavior. Use a read-only or test config and tiny amounts (or testnet) first. - Missing declared dependencies: The logging script uses jq but jq is not declared. Ensure required binaries are installed and consider running in a sandbox before granting network/wallet access. - Provenance and author: The repository/author metadata is unknown and the README references an odd username. Consider source provenance and prefer well-known/trusted integrations for money-handling skills. - If you proceed: (1) audit the Bankr config contents, (2) run the skill in sandbox/testnet, (3) restrict autonomous execution until you confirm safe behavior, and (4) keep small trade sizes and revoke keys if anything suspicious happens.

Autonomous trading system for Base chain via Bankr API.

Philosophy

"The market is a machine for transferring wealth from the impatient to the patient."

This skill prioritizes capital preservation over aggressive gains. Every trade has defined risk parameters. No YOLO.

Prerequisites

• Bankr API configured at ~/.clawdbot/skills/bankr/config.json
• ETH in your Bankr wallet for gas and trading
• Check your wallet address with: bankr.sh "What is my wallet address?"

Core Principles

1. Risk Management (NON-NEGOTIABLE)

MAX_POSITION_SIZE = 10% of portfolio
MAX_SINGLE_TRADE = $50 USD
STOP_LOSS = -15% from entry
TAKE_PROFIT = +30% minimum target
MAX_DAILY_LOSS = -20% of starting balance

If daily loss limit hit → STOP TRADING FOR 24 HOURS.

2. Entry Criteria (ALL must pass)

Before ANY buy:

• Liquidity > $10k
• Contract verified on Basescan
• No honeypot indicators (can sell)
• Not on known scam list
• Age > 5 minutes (avoid rugs at launch)
• Price action shows accumulation, not dump

3. Exit Rules

Take Profit (scale out):

• 25% at +30%
• 25% at +50%
• 25% at +100%
• Hold 25% moonbag

Stop Loss:

• Hard stop at -15%
• No averaging down on losers

Trading Strategies

Strategy 1: Launch Sniper (HIGH RISK)

Monitor new token launches, enter early with small size.

# Check trending/new tokens
scripts/bankr.sh "What tokens are trending on Base?"

# Research before buying
scripts/bankr.sh "What's the liquidity for TOKEN on Base?"

# Small entry
scripts/bankr.sh "Buy $20 of TOKEN on Base"

Rules:

• Max $20 per snipe
• Sell 50% at 2x, rest at 5x or stop loss
• Max 3 snipes per day

Strategy 2: Momentum Trading (MEDIUM RISK)

Trade established tokens showing strength.

Entry signals:

• Price > 20-period MA
• Volume increasing
• Higher lows forming

# Check momentum
scripts/bankr.sh "Do technical analysis on TOKEN"

# Enter with limit order
scripts/bankr.sh "Buy $30 of TOKEN if price drops to X"

Strategy 3: DCA Blue Chips (LOW RISK)

Steady accumulation of proven tokens.

# Weekly DCA
scripts/bankr.sh "DCA $20 into ETH every week on Base"
scripts/bankr.sh "DCA $10 into USDC every week on Base"

Execution via Bankr

Check Portfolio

~/clawd/skills/bankr/scripts/bankr.sh "Show my portfolio on Base"

Execute Trade

~/clawd/skills/bankr/scripts/bankr.sh "Buy $25 of TOKEN on Base"

Set Stop Loss

~/clawd/skills/bankr/scripts/bankr.sh "Set stop loss for TOKEN at -15%"

Check Price

~/clawd/skills/bankr/scripts/bankr.sh "What's the price of TOKEN on Base?"

Trade Journal

Log every trade to data/trades.json:

{
  "timestamp": "2026-01-28T12:00:00Z",
  "action": "BUY",
  "token": "TOKEN",
  "amount_usd": 25,
  "price": 0.001,
  "reason": "Launch snipe - verified contract, good liquidity",
  "tx": "0x..."
}

After each trade, update the journal. Review weekly for pattern analysis.

Daily Routine

Morning (9 AM)

1. Check portfolio balance
2. Review overnight price action
3. Identify opportunities
4. Set limit orders for the day

Midday (1 PM)

1. Check open positions
2. Adjust stop losses if in profit
3. Take profits if targets hit

Evening (6 PM)

1. Close any day trades
2. Log all trades to journal
3. Calculate daily PnL
4. Review what worked/didn't

Red Flags (DO NOT TRADE)

• Honeypot (can't sell)
• Liquidity < $5k
• Unverified contract
• Team anonymous with no track record
• Promises of guaranteed returns
• Excessive tax (>10%)
• Locked liquidity < 30 days
• Price already 10x+ from launch

Performance Tracking

Track in data/performance.json:

{
  "start_date": "2026-01-28",
  "starting_balance_usd": 100,
  "current_balance_usd": 100,
  "total_trades": 0,
  "winning_trades": 0,
  "losing_trades": 0,
  "win_rate": 0,
  "total_pnl_usd": 0,
  "best_trade": null,
  "worst_trade": null
}

Safety Overrides

If ANY of these occur, STOP ALL TRADING:

1. Daily loss > 20%
2. 3 consecutive losing trades
3. Portfolio down > 30% from ATH
4. Unexpected error in execution
5. Market-wide crash (ETH -20% in 24h)

Wait 24 hours, reassess, then resume with smaller size.

Autonomous Trading Mode

When running autonomously (via cron or heartbeat):

Morning Scan (9 AM)

1. Check portfolio balance
2. Review overnight price action on holdings
3. Scan for new opportunities
4. Set limit orders for the day

Midday Check (1 PM)

1. Monitor open positions
2. Adjust trailing stops on winners
3. Take profits if targets hit
4. Log any executed trades

Evening Review (6 PM)

1. Close day trades if any
2. Calculate daily PnL
3. Update performance.json
4. Generate summary for user

Execution Commands

# Morning
~/clawd/skills/bankr/scripts/bankr.sh "Show my portfolio on Base"
~/clawd/skills/bankr/scripts/bankr.sh "What tokens are trending on Base?"

# Execute trade
~/clawd/skills/bankr/scripts/bankr.sh "Buy $25 of TOKEN on Base"

# Set protection
~/clawd/skills/bankr/scripts/bankr.sh "Set stop loss for TOKEN at -15%"

# Take profit
~/clawd/skills/bankr/scripts/bankr.sh "Sell 25% of my TOKEN on Base"

References

Core Trading

• references/strategies.md - Detailed strategy breakdowns
• references/token-analysis.md - How to analyze tokens
• references/risk-management.md - Position sizing formulas

Market & Research

• references/market-analysis.md - Reading market conditions
• references/market-research-bankr.md - Bankr research commands

Execution

• references/execution.md - Order types and execution
• references/automation-strategies.md - Automated trading setups
• references/launch-sniping.md - New token launch guide

Advanced

• references/leverage-guide.md - Leveraged trading (high risk)

---

"The goal is not to make money on every trade. The goal is to be profitable over time."