adaptive-playbook

Operational Implementation of the Adaptive Interfaces Guide

Adaptive Interfaces
Engineering discipline for configuring agents to produce team-conforming output.

1. Purpose

This repository defines the execution protocol for configuring agents so their output:

• matches team structure and conventions
• aligns with domain expectations
• passes review without structural correction

This repository is operational, not conceptual.

2. Position in the Framework

This repository implements the HOW.

| Layer | Repository | Role | |||| | ACS | adaptive-conformance-specification | behavior discipline | | ATD | adaptive-tool-discovery | tool capability mapping | | AO | adaptive-onboarding | team conventions | | GUIDE | adaptive-guide | conceptual framework | | PLAYBOOK | adaptive-playbook | execution protocol |

3. Scope: Included

3.1 Repository Core Artifacts

• PLAYBOOK.md — step-by-step execution process
• QUICKSTART.md — minimal setup for engineers
• CHECKLIST.md — evaluation and PR enforcement
• ANNOTATIONS.md — supporting notes

3.2 Repository Function

Defines:

• how to establish priors
• how to load context layers
• how to execute tasks with an agent
• how to evaluate and iterate on output

4. Scope: Not Included

• behavioral protocol: see ACS
• tool registries: see ATD
• team context schemas: see AO
• conceptual explanation: see GUIDE
• domain implementations: see example repos

5. Minimal Workflow

1. Generate AO context from a repository
2. Review and correct inferred conventions
3. Define domain context (required for correctness)
4. Load:
   • ACS
   • AO config
   • AO context
   • domain context
5. Execute task
6. Evaluate output
7. Update priors and repeat

6. Core Principle

Agent output quality is determined by configuration, not prompting.

Prompts are task instructions; priors are the system.

7. Developer

npx markdownlint-cli@latest "**/*.md"

8. Annotations

ANNOTATIONS.md

9. Manifest

MANIFEST.toml

10. Citation

CITATION.cff

11. License

MIT