agent-conduct

Scan changed files for hardcoded agent names and verify tests before merge.

Updated Mar 18, 2022
One-click install
npx skills add https://github.com/mpaulosky/dotfiles --skill agent-conduct-mpaulosky
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-conduct
Source: https://github.com/mpaulosky/dotfiles/tree/main/.copilot/skills/agent-conduct
Command: npx skills add https://github.com/mpaulosky/dotfiles --skill agent-conduct-mpaulosky

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill centralizes and enforces two hard-team rules that were previously duplicated across many agent charters: preventing hardcoded agent names from seeding product code and ensuring contributors run tests for files they changed. Centralizing these rules reduces accidental test breakage, brittle CI configurations, and cross-squad coupling in product logic.

Core Features & Use Cases

  • Product Isolation Rule: Detect and forbid hardcoded squad/agent names in tests, CI configs, and product code, recommending parameterized or clearly fake test fixtures instead.
  • Peer Quality Check: Require verification that the test suite for modified files runs locally or in CI before declaring work done and updating history.md when mistakes are learned.
  • Use Case: When a developer opens a PR that references a specific agent name in an assertion or workflow, run checks to replace it with a test fixture and confirm affected tests pass so the PR does not worsen pre-existing CI failures.

Quick Start

Apply the agent-conduct rules by scanning changed files for literal agent names and verifying the relevant tests run successfully before completing the pull request.

Frequently Asked Questions about agent-conduct

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent hardcoded agent names from breaking CI workflows?

Preventing hardcoded agent names from breaking CI workflows requires scanning changed files and CI configurations for literal agent identifiers. Replace these references with parameterized or clearly fake test fixtures to ensure product isolation.

What is a product isolation rule for agent governance in tests?

A product isolation rule for agent governance detects and forbids hardcoded squad or agent names in tests, CI configs, and product code. It recommends parameterized test fixtures instead to reduce accidental test breakage and cross-squad coupling.

How do I verify tests pass for changed files before merging a pull request?

Verifying tests pass for changed files before merging a pull request requires running the relevant test suite locally or in CI. Contributors must confirm successful test execution for modified files prior to declaring work done and updating history.md.

Why does referencing a specific agent name in an assertion cause test failures?

Referencing a specific agent name in an assertion causes test failures by creating brittle tests and cross-squad coupling. Replacing the literal agent identifier with a parameterized test fixture prevents worsening pre-existing CI failures.

Can I use a peer quality check to enforce local test execution for modified files?

Yes, you can use a peer quality check to enforce local test execution for modified files. It requires contributors to verify that the test suite for changed files runs successfully locally or in CI before declaring work done.

When should I not use hardcoded agent identifiers in CI configuration?

You should not use hardcoded agent identifiers in CI configuration when tests or workflows reference specific squads. Parameterized test fixtures prevent brittle CI configurations and accidental test breakage from cross-squad coupling.