| name | continuous-learning-pattern-extraction |
| description | Extract and learn patterns from trading data continuously. Use when identifying recurring patterns, building adaptive models, or continuous improvement workflows.
|
| version | 1.0.0 |
| allowed-tools | Read, Write, Edit, Bash |
Continuous Learning Pattern Extraction
Source: https://mcpmarket.com/tools/skills/continuous-learning-pattern-extraction
Use When
- You need to apply this capability as part of trading research workflows (data, features, backtests, ML, reporting).
- You want a reproducible output that can be committed to this repo (code, configs, docs).
Inputs To Ask For
- Objective: what success looks like (metric, constraints, time horizon).
- Data: symbols, timeframe, sampling, data sources, and leakage risks.
- Constraints: compute budget, latency, interpretability, and deployment requirements.
Outputs
- A concrete plan (steps + checks).
- A minimal implementation sketch (files to create/change) and verification steps.
- If applicable: a risk checklist (leakage, overfitting, evaluation pitfalls).
Workflow
- Restate the task in measurable terms.
- Enumerate required artifacts (datasets, features, configs, scripts, reports).
- Propose a default approach and 1-2 alternatives.
- Add validation gates (walk-forward, Monte Carlo, sanity checks).
- Produce repo-ready deliverables (code + docs) and a run command.
