mcpField
W
whale-index

whale-index

github.com/senpi-aiOther140141 installsUpdated today
Repository

About whale-index

whale-index is an automated trader-mirroring skill for traders who want portfolio selection and daily oversight based on a budget and risk level. It scores Discovery leaderboard candidates, requests approval before creating strategies, and applies defined monitoring and replacement rules.

Key features

  1. 01Filter traders by budget and risk
  2. 02Score candidates and check position overlap
  3. 03Weight allocations and set stop losses
  4. 04Monitor daily and replace by defined criteria
  5. 05Close strategies and return funds

Use cases

  1. 01Build a mirror portfolio for a chosen risk level
  2. 02Review strategy performance and rebalance daily
  3. 03Assess replacements when traders deteriorate

MCP support

The skill depends on APIs for trader discovery, mirror strategy creation, position checks, and strategy closure. The README does not specify how these APIs are exposed through MCP.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/senpi-ai/senpi-skills --skill whale-index

After installing, type/whale-indexin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

readme.md
references/
daily-monitoring.md
skill-attribution.md

Details

From the project README

Whale Index — Auto-Mirror Top Discovery Traders

Scan the Discovery leaderboard, score traders, create mirror strategies, and rebalance daily. Set your risk level and budget — the agent handles selection, allocation, monitoring, and swaps.

Skill Attribution

When creating a strategy, include skill_name and skill_version in the call. See references/skill-attribution.md for details.

5-Step Flow

Step 1: Onboard the User

Collect: budget, risk tolerance (conservative/moderate/aggressive).

BudgetSlots
$500-$2k2
$2k-$5k3
$5k-$10k4
$10k+5

Risk mapping:

RiskAllowed LabelsMax Leverage
ConservativeELITE only10x
ModerateELITE, RELIABLE15x
AggressiveELITE, RELIABLE, BALANCED25x
Step 2: Discover Traders

2a. Pull candidates: discovery_top_traders(limit=50, timeframe="30d")

2b. Hard filters:

  • Consistency label matches risk level
  • Risk label matches risk level
  • Min 30d track record
  • Not already in user's portfolio

2c. Score remaining candidates:

score = 0.35 × pnl_rank + 0.25 × win_rate + 0.20 × consistency + 0.10 × hold_time + 0.10 × drawdown

All components normalized 0-100.

2d. Overlap check: Compare active positions across selected traders. Flag >50% position overlap.

2e. Allocation weighting: Score-weighted allocation with 35% cap per slot. Re-normalize after capping.

Step 3: Present & Confirm

Show the user: trader address, rank, labels, win rate, allocation amount. Wait for approval before executing.

Step 4: Execute

For each slot:

  1. Create mirror strategy via strategy_create_mirror
  2. Set strategy-level stop loss (-10% conservative, -15% moderate, -25% aggressive)
  3. Confirm mirroring is active
Step 5: Daily Monitoring (Cron)

See references/daily-monitoring.md for the complete daily check procedure, swap criteria, and rebalance logic.

Swap criteria (ALL must be true):

  1. Degraded: dropped below rank 50 OR consistency fell OR inactive 48h+ OR drawdown 2× historical
  2. Sustained: WATCH status for 2+ consecutive days (tracked via watchCount)
  3. Better alternative: replacement scores ≥15% higher
  4. User's strategy-level SL not hit

Key principle: The 2-day watch period prevents churn from temporary dips.

Teardown

To exit: close all mirror strategies, return funds to main wallet.

API Dependencies

  • discovery_top_traders — trader leaderboard
  • strategy_create_mirror — create mirror strategy
  • strategy_get_clearinghouse_state — check positions
  • strategy_close_strategy — teardown

Fee Estimates

Mirror strategies incur the same trading fees as the mirrored trader's activity. Budget ~0.5-1% daily in fees for active traders.

Comments

Markdown supported