Overfitting Detector — Development agent for Claude Code
Adversarial reviewer that tries to disprove a strategy's backtest.
How to install Overfitting Detector
Installs to ~/.claude/agents/abhi15724-quantforge-ai-overfitting-detector.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/abhi15724/quantforge-ai/HEAD/agents/overfitting-detector.md -o ~/.claude/agents/abhi15724-quantforge-ai-overfitting-detector.md Restart Claude Code, or start a new session, for it to be picked up.
What Overfitting Detector does
name: overfitting-detector description: Adversarial reviewer that tries to disprove a strategy's backtest. Use after any backtest or when a result looks too good.
You are the Overfitting Detection agent. Your job is to disprove the result, not to be polite. Check look-ahead, survivorship, leakage, selection bias, multiple testing, parameter count vs trades, parameter sensitivity, unrealistic fills, cost sensitivity, regime dependence, and time-stability. For each, answer PASS / FAIL / U
Alternatives in Development
- Strategy Reviewer 1k ★
- Dast Devils Advocate — Adversarial validator for DAST findings 812 ★
- Finding Checker — Blind adversarial checker — given ONLY a finding artifact and its evidence (never the author's reasoning), tri 348 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Finding Validator
Adversarial validation agent that attempts to disprove blocking findings from code review. Receives a finding
Reviewer Refute
Refute phase of deep-review — tries to disprove one dimension's findings against the actual code. Spawned by d
Adversarial Coach
Adversarial code reviewer that tries to break code. Use immediately after any significant code change to find
Breaker
Adversarial code reviewer that tries to break code. Use immediately after any significant code change to find
Oratores Critic
Independent reviewer for persuasive work already produced — a speech draft, a deck, a chart, an audience strat
Strategy Critic
PROACTIVELY use this subagent immediately after the user writes or revises any strategy artifact: Strategy Blo
Related Skills
QA Strike Team
An adversarial QA and security review skill for apps you build with Claude Code. You point it at a running app
Desk Test
Backtest a structure with modelled premiums and judge honestly whether the result means anything
New Strategy
Create a new STRATEGY.md from an idea, strategy code, or a backtest, validate it, and start it in paper status