Counterfactual Statistics — AI skill for Claude Code
Statistics for counterfactual and paired-branch ML experiments - the independent unit, paired tests against a no-op anchor, grouped splits, winner's curse, abstention calibration, pre-registration, an.
How to install Counterfactual Statistics
Installs to ~/.claude/commands/tachyon-beep-skillpacks-counterfactual-statistics.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/tachyon-beep/skillpacks/HEAD/.claude/commands/counterfactual-statistics.md -o ~/.claude/commands/tachyon-beep-skillpacks-counterfactual-statistics.md Restart Claude Code, or start a new session, for it to be picked up.
What Counterfactual Statistics does
description: Statistics for counterfactual and paired-branch ML experiments - the independent unit, paired tests against a no-op anchor, grouped splits, winner's curse, abstention calibration, pre-registration, and reliability reporting
Counterfactual Statistics Routing
**The independent unit is the run you forked, not the branch you forked it into.** Branches bought you precision, not sample size — nearly every failure in this domain follows from forgetting that, and the rest from c
Alternatives in AI
- Ccstatusline — by sirmalloc - A highly customizable status line formatter for Claude Code CLI that displays model info, git b 5.5k ★
- Riteway — Simple, readable, helpful unit tests 1.2k ★
- Rmslop — Remove AI code slop from current branch 508 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Quantum Calibration Agent Blueprint
This is a reference agent blueprint for AI-powered quantum device calibration. It provides an intelligent agen
Beacon Skill Rs
Rust crate for Beacon Atlas relay — AI agent registration, heartbeat, SEO-enhanced discoverability. Published
Skillranker
Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context.
Sliced Review
Claude Code skill that has other AI models (GPT, Grok, GLM, and more) review your whole codebase for bugs. It
Elop Crew
ELOP Crew ? the AI Agent Monitor for Claude Code: a local dashboard and desktop app for every Claude Code sess
Iterm2 Agent Tabs
Turn iTerm2 into a workbench for running many Claude Code / Codex sessions: live state icons and tab colors dr
Related Agents
Librarian Critic
Literature quality critic. Reviews the Librarian's annotated bibliography for coverage gaps, journal quality,
Reform Comparator
Stage-5 comparator. Takes the microsim output and the prior-scores anchor, returns PASS (within expected band,
Planted Error Probe
The ungameable calibration anchor — inject a KNOWN bug into a critical module and confirm the test suite catch