feat(chatmode): add GPT-5 Beast Mode high-agency engineering assistant

[nicholasdbrady]

Adds a new chat mode: GPT-5 Beast Mode.

Purpose:

• High-agency autonomous engineering persona for complex multi-step tasks
• Emphasizes verification (tests/linters), minimal-risk iterative changes, and decisive progress
• Useful for refactors, debugging, architecture navigation, automation, and multi-repo tasks

Why it’s distinct:

• Enforces a strict tool preamble & disciplined workflow
• Explicit stop conditions and destructive action planning requirement
• High-signal, no-filler communication with strong guardrails
• Integrates planning + validation loops tightly

Attribution:
Submitted by GitHub user: nicholasbrady
Authored collaboratively with the GPT-5 Beast Mode persona itself

Checklist:

• File added under chatmodes/ with correct naming & extension
• Frontmatter includes description, model, tools, tags, attribution
• Persona is specific and non-duplicative
• Tested locally for role persistence and constraint adherence
• Conforms to contribution guidelines

Follow-up (optional):

• Add collection bundling refactor/test triage prompts + this mode
• Provide sample session in a collection README

Thanks!

[Copilot]

Pull Request Overview

This PR introduces a specialized chat mode called "GPT-5 Beast Mode" designed as a high-agency autonomous engineering assistant for complex, multi-step software development tasks. The chat mode emphasizes tool-driven workflows, verification-first approaches, and disciplined iterative development.

Key changes:

• Adds a comprehensive engineering assistant persona with strict operational guidelines
• Implements a structured workflow with explicit tool usage preambles and stop conditions
• Defines clear communication patterns prioritizing concrete actions over speculation

[Copilot]

Pull Request Overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

[aaronpowell]

Can you run the update script to update the readme.

Also, I'd suggest either adding a new level one heading (single #) at the start of the file or a title to the front matter to ensure an accurate name in the README.

[aaronpowell]

For some reason it seems like the update and validation scripts aren't working. Can you allow contributors to edit the PR and I'll fix it