ce-reject-policy

Community

Policy-driven rejection for model predictions.

AuthorMoffran
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Configures and integrates reject and defer decision policies into production prediction and explanation flows, addressing uncertainty by optionally skipping, flagging, or deferring cases.

Core Features & Use Cases

  • Implements four policies: NONE, FLAG, ONLY_REJECTED, ONLY_ACCEPTED, plus deprecation-mapped legacy names.
  • Provides a RejectResult envelope exposing prediction, explanation, rejection mask, policy, and telemetry metadata.
  • Supports per-call policy overrides and explainer-level defaults, enabling flexible governance in live systems.
  • Demonstrates integration patterns with calibration outcomes and how to inspect init_error and rejection metrics.

Quick Start

Run an end-to-end example using the FLAG policy to flag uncertain predictions and inspect the RejectResult metadata.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: ce-reject-policy
Download link: https://github.com/Moffran/calibrated_explanations/archive/main.zip#ce-reject-policy

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.