Does this skill work with any repository?

While highly effective generally, it is optimized for dev stacks using the 'Beads' state management and 'North Star' documentation patterns found in the Knowledge and Vibes repository.

What is 'tests over rhetoric'?

This is the philosophy that technical truth is found in code execution. Instead of arguing which approach is better, the skill writes tests to prove which implementation actually meets the requirements.

When should I use the calibrate skill?

Run the calibrate skill between project phases, when agents disagree on a technical approach, or whenever you suspect the code is drifting away from the original project requirements.

How does the orchestrator-subagent pattern benefit development?

It runs each analysis phase in a fresh, isolated context. This prevents 'lost in the middle' errors and performance degradation that often occurs in long-context AI conversations.

Calibration Orchestrator

Name: Calibration Orchestrator
Author: Mburdo

byMburdo

•

Productivity & Workflow

Synchronizes AI development workflows with core project goals using evidence-based subagent audits and drift detection.

About

The Calibrate skill implements a sophisticated multi-agent orchestration pattern designed to prevent project drift and ensure strict adherence to a project's 'North Star' requirements. By spawning specialized subagents in fresh execution contexts, it performs deep analysis across coverage analysis, drift detection, and test-based challenges. This skill replaces subjective debate with objective evidence, adjudicating technical disagreements through discriminating tests and generating formal synthesis reports to keep complex AI-driven development tasks on track and aligned with user expectations.

Key Features

Evidence-Based Calibration: Prioritizes verifiable test results and code-level evidence over rhetorical debate or AI persuasion.
Orchestrator-Subagent Pattern: Spawns isolated agents for coverage, drift, and synthesis to eliminate context degradation.
Automated Drift Detection: Automatically identifies deviations between current progress and the project's established North Star goals.
46 GitHub stars
Discriminating Testing: Resolves architectural disagreements by executing tests designed to pass for one approach and fail for another.
Synthesis Reporting: Aggregates findings from multiple audits into actionable user reports that preserve dissenting opinions.

Use Cases

Resolving logic or architecture disagreements between AI agents through objective test-based adjudication.
Realigning technical execution with project requirements after completing major development phases.
Conducting deep-dive coverage and security audits when the user detects project drift or scope creep.

About

Key Features

Evidence-Based Calibration: Prioritizes verifiable test results and code-level evidence over rhetorical debate or AI persuasion.
Orchestrator-Subagent Pattern: Spawns isolated agents for coverage, drift, and synthesis to eliminate context degradation.
Automated Drift Detection: Automatically identifies deviations between current progress and the project's established North Star goals.
46 GitHub stars
Discriminating Testing: Resolves architectural disagreements by executing tests designed to pass for one approach and fail for another.
Synthesis Reporting: Aggregates findings from multiple audits into actionable user reports that preserve dissenting opinions.

Use Cases

Resolving logic or architecture disagreements between AI agents through objective test-based adjudication.
Realigning technical execution with project requirements after completing major development phases.
Conducting deep-dive coverage and security audits when the user detects project drift or scope creep.