Name: my-awesome-copilot
Author: JakeTRogers

Imported from JakeTRogers/my-awesome-copilot (AGENTS.md). Install upstream with npx skills add JakeTRogers/my-awesome-copilot. Copyright stays with the author.

my-awesome-copilot

Personal collection of GitHub Copilot / Claude Code prompts, agents, instructions, and skills — a public notebook of patterns actually in use, not an awesome-list fork.

Commands

Validate everything: pre-commit run --all-files (large files, EOF, merge conflicts, JSON/YAML, trailing whitespace, commit-msg format)
There is no build or test suite; the artifacts are markdown files consumed by AI tooling.

Architecture

skills/ — portable Agent Skills (<name>/SKILL.md) usable by both Copilot and Claude Code, e.g. /pull-request. Authoring guide: .github/instructions/agent-skills.instructions.md.
agents/ — a Maestro-orchestrated multi-agent system (*.agent.md): Architect, Auditor, Chronicler, Coder, Researcher, Tester, plus the standalone Promptly prompt-engineering mode. Authoring guide: .github/instructions/agents.instructions.md.
instructions/ — language/format conventions applied by glob (applyTo frontmatter): Go, Python, shell, markdown.

Conventions

Commits: Conventional Commits enforced by commitizen (.cz.yaml) + pre-commit commit-msg hook. Never use the chore type; never use multiple scopes. Commits are GPG-signed by the user via YubiKey — agents must never create commits or tags; hand the user the commands instead.
Markdown: follow instructions/markdown.instructions.md — blank lines around headings/lists/fences, language on every fence, no bare URLs.
Skill frontmatter must stay portable across Copilot and Claude Code (agentskills.io spec); descriptions must state WHAT the skill does and WHEN to use it.
Prompts are self-contained (they must work where personal skills are not installed) but their rules must stay content-aligned with the corresponding skill.

Gotchas

Skill bodies embedding markdown templates that contain code fences must use 4-backtick outer fences to avoid premature fence termination. Otherwise use standard 3-backtick fences.

{ "ocm": "1", "id": "jaketrogers-my-awesome-copilot-my-awesome-copilot", "kind": "skill", "name": "my-awesome-copilot", "description": "Personal collection of GitHub Copilot / Claude Code prompts, agents, instructions, and skills — a public notebook of patterns actually in use, not an awesome-list fork.", "publisher": "JakeTRogers", "version": "1.0.0", "capabilities": { "domains": [ "coding" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Personal collection of GitHub Copilot / Claude Code prompts, agents, instructions, and skills — a public notebook of patterns actually in use, not an awesome-list fork." ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/JakeTRogers/my-awesome-copilot", "path": "AGENTS.md", "ref": "2a7f4b4c24b3f545f75c0633fc414bcce6f8cd61", "url": "https://github.com/JakeTRogers/my-awesome-copilot/blob/2a7f4b4c24b3f545f75c0633fc414bcce6f8cd61/AGENTS.md", "key": "JakeTRogers/my-awesome-copilot/AGENTS.md" } }, "instructions": "# my-awesome-copilot\n\nPersonal collection of GitHub Copilot / Claude Code prompts, agents, instructions, and skills — a public notebook of patterns actually in use, not an awesome-list fork.\n\n## Commands\n\n- Validate everything: `pre-commit run --all-files` (large files, EOF, merge conflicts, JSON/YAML, trailing whitespace, commit-msg format)\n- There is no build or test suite; the artifacts are markdown files consumed by AI tooling.\n\n## Architecture\n\n- `skills/` — portable Agent Skills (`<name>/SKILL.md`) usable by both Copilot and Claude Code, e.g. `/pull-request`. Authoring guide: `.github/in", "cost": { "context_tokens": 489 } }

my-awesome-copilot

About