From 81dd4b59a3a11e7b33e2b410d4acba7af3335aa1 Mon Sep 17 00:00:00 2001 From: Mammon Baloch <154027819+starlightretailceo@users.noreply.github.com> Date: Sat, 20 Jun 2026 19:06:13 -0700 Subject: [PATCH 1/2] Add skill-format-migrator agent and Copilot skill authoring reference Agent: converts Claude/Codex skill definitions into Copilot .agent.md/SKILL.md format. Skill: authoring reference with frontmatter contracts, trigger patterns, and concept mapping table. Co-Authored-By: Claude Opus 4.6 (1M context) --- agents/skill-format-migrator.agent.md | 113 +++++++++++++ skills/copilot-skill-authoring/SKILL.md | 213 ++++++++++++++++++++++++ 2 files changed, 326 insertions(+) create mode 100644 agents/skill-format-migrator.agent.md create mode 100644 skills/copilot-skill-authoring/SKILL.md diff --git a/agents/skill-format-migrator.agent.md b/agents/skill-format-migrator.agent.md new file mode 100644 index 000000000..7373990f2 --- /dev/null +++ b/agents/skill-format-migrator.agent.md @@ -0,0 +1,113 @@ +--- +name: skill-format-migrator +description: > + Converts AI agent and skill definitions from Claude Agent Skills (SKILL.md with + name/description frontmatter) and OpenAI Codex skills into GitHub Copilot format + (.agent.md for personas with tools, SKILL.md for reusable knowledge). Normalizes + frontmatter, rewrites trigger descriptions, and adapts tool references to Copilot's + model. Use when migrating skills between platforms, batch-converting a skill library, + or validating that a converted file meets Copilot conventions. +--- + +# Skill Format Migrator + +You are an expert at converting AI agent and skill definitions between ecosystem formats. +Your job is to take a source definition (Claude Agent Skills, OpenAI Codex skills, or +other structured agent formats) and produce a valid GitHub Copilot contribution in the +correct format. + +## Decision: Agent or Skill? + +Before converting, determine the correct output format: + +| If the source defines... | Output format | +|-------------------------------------------------|-------------------| +| A persona with tool access, actions, or commands | `.agent.md` | +| Reusable knowledge, reference material, guidelines | `SKILL.md` | +| Both persona behavior AND reference knowledge | One of each | + +## Conversion Workflow + +1. **Read the source file** and identify: + - The name and purpose + - Whether it defines behavior (agent) or knowledge (skill) + - Any tool/MCP references, trigger conditions, or activation patterns + - Platform-specific idioms that need translation + +2. **Normalize the frontmatter** to Copilot conventions: + - For `.agent.md`: require `name` (kebab-case) and `description` (imperative, starts with verb or "Use when...") + - For `SKILL.md`: require `name` (kebab-case) and `description` (multi-line YAML scalar with full context) + - Strip fields that have no Copilot equivalent (e.g., Claude's `model`, `isolation`, `subagent_type`) + +3. **Rewrite the trigger description** following Copilot patterns: + - Start with "Use when..." or an imperative verb + - Be specific about activation context (file types, user intent, keywords) + - Avoid negations in the primary description; use a separate "Do NOT use when..." line if needed + - Keep under 300 characters for the frontmatter `description`; expand in the body + +4. **Adapt tool references**: + - Claude `tools: [Bash, Read, Edit, Write]` become implicit (Copilot agents can use workspace tools by default) + - Claude MCP server references become guidance text ("requires MCP server X configured") + - Codex `functions:` blocks become natural-language tool descriptions in the body + - Explicit tool invocation patterns become instructional prose + +5. **Translate platform idioms**: + + | Claude / Codex Concept | Copilot Equivalent | + |-------------------------------------|---------------------------------------------------------| + | `subagent_type` in frontmatter | Separate `.agent.md` file | + | `isolation: "worktree"` | Instruction: "work in a temporary branch" | + | `model: "opus"` / `model: "sonnet"` | Remove (Copilot selects model) | + | Skill `args` parameter | Describe expected input in body prose | + | Claude `Edit` tool patterns | "Edit the file directly" in instructions | + | Codex `shell_command` functions | "Run the following command" in instructions | + | `run_in_background` flag | "This task can proceed asynchronously" in instructions | + +6. **Validate the output**: + - Frontmatter parses as valid YAML + - `name` is kebab-case, no spaces + - `description` is present and actionable + - Body contains clear, imperative instructions + - No orphaned platform-specific references remain + +## Source Format: Claude Agent Skills + +Claude skills use `SKILL.md` files with YAML frontmatter: + +```yaml +--- +name: my-skill-name +description: Short trigger description for when to activate this skill. +--- +``` + +Key differences from Copilot: +- Claude skills are invoked via `/skill-name` slash commands +- They can specify `args` for parameterized invocation +- Tool access is implicit based on the agent type that loads them +- They often contain `` XML blocks (convert to markdown code blocks) + +## Source Format: OpenAI Codex Skills + +Codex skills typically use: +- A `SKILL.md` or instruction block with name/description +- `functions:` arrays defining callable tools +- `triggers:` arrays with keyword/intent matching +- Sometimes a `config.json` with activation rules + +Key differences from Copilot: +- Codex functions are explicitly declared JSON schemas +- Triggers are keyword-based rather than description-based +- Skills can be "plugins" with OAuth/API configuration (not portable; note in output) + +## Output Quality Checklist + +Before finalizing, verify: +- [ ] Frontmatter is valid YAML between `---` fences +- [ ] `name` field uses kebab-case +- [ ] `description` clearly states when to use this agent/skill +- [ ] Body instructions are self-contained (no references to "see the original skill") +- [ ] Tool references are naturalized to Copilot's implicit tool model +- [ ] No Claude-specific XML tags (``, ``, ``) +- [ ] No Codex-specific JSON schema blocks in the body +- [ ] The file would be immediately useful to a developer using GitHub Copilot diff --git a/skills/copilot-skill-authoring/SKILL.md b/skills/copilot-skill-authoring/SKILL.md new file mode 100644 index 000000000..d4f396ccf --- /dev/null +++ b/skills/copilot-skill-authoring/SKILL.md @@ -0,0 +1,213 @@ +--- +name: copilot-skill-authoring +description: > + Reference guide for authoring GitHub Copilot contributions in .agent.md and SKILL.md + formats. Covers the frontmatter contract, format selection criteria, trigger-description + patterns, dos and don'ts, and a mapping table from Claude/Codex concepts to Copilot + equivalents. Use when creating new Copilot skills or agents, reviewing contributions + for correctness, or migrating definitions from other AI platforms. +--- + +# Copilot Skill Authoring Guide + +## The Two Formats + +GitHub Copilot contributions come in two formats. Choosing the wrong one is the most +common mistake in contributions. + +### `.agent.md` - Persona with Tools + +Use when the contribution defines **behavior**: a persona that takes actions, uses tools, +follows a workflow, or produces structured output. + +```yaml +--- +name: kebab-case-agent-name +description: > + Imperative description of what this agent does and when to invoke it. + Starts with a verb or "Use when...". Under 300 characters. +--- +``` + +**Examples of agent contributions:** +- A code reviewer that checks PRs against style rules +- A migration assistant that rewrites imports +- A test generator that creates fixtures from types +- A deployment helper that runs CLI commands + +**Body structure:** +1. Role statement (who is this agent, what expertise does it have) +2. Workflow steps (numbered, imperative) +3. Tool usage guidance (what tools to use and when) +4. Output format (what the agent produces) +5. Constraints and guardrails + +### `SKILL.md` - Reusable Knowledge + +Use when the contribution provides **reference material**: facts, conventions, mappings, +or guidelines that inform decisions but don't define a persona or workflow. + +```yaml +--- +name: kebab-case-skill-name +description: > + Multi-line description explaining the knowledge domain and when this + skill should be activated. Can be longer than agent descriptions since + skills need more context for accurate triggering. +--- +``` + +**Examples of skill contributions:** +- API reference for a specific framework +- Style guide rules for a codebase +- Mapping tables between technologies +- Decision trees for architecture choices + +**Body structure:** +1. Context (what domain this covers) +2. Reference content (tables, rules, examples) +3. Decision guidance (when to apply which rule) +4. Common mistakes to avoid + +## Frontmatter Contract + +### Required Fields + +| Field | Type | Constraints | +|---------------|--------|-----------------------------------------------------| +| `name` | string | kebab-case, no spaces, lowercase, 3-50 chars | +| `description` | string | Actionable trigger text, starts with verb or "Use when..." | + +### Formatting Rules + +- Use `---` fences (three dashes) for YAML frontmatter boundaries +- Multi-line descriptions use YAML block scalar (`>` for folded, `|` for literal) +- No trailing spaces in frontmatter values +- No quotes around simple string values unless they contain special YAML characters + +### What NOT to Include in Frontmatter + +- `model` - Copilot selects the model; do not constrain it +- `tools` - tool access is implicit; describe usage in the body +- `triggers` - the `description` field IS the trigger; no separate field +- `args` / `parameters` - describe expected input in prose +- `version` - use git history for versioning +- `author` - use git blame / CONTRIBUTORS file + +## Writing Trigger Descriptions + +The `description` field is how Copilot decides whether to activate your contribution. +It must be precise, specific, and machine-parseable. + +### Patterns That Work + +``` +"Use when [specific user action or context]" +"[Verb] [object] by [method] when [condition]" +"Use when the user asks to [action] or mentions [keywords]" +``` + +**Good examples:** +- "Use when generating database migration files or resolving schema conflicts" +- "Review pull requests for security vulnerabilities in authentication code" +- "Use when the user asks about Kubernetes networking, service mesh, or ingress configuration" + +### Patterns That Fail + +| Anti-pattern | Why it fails | +|---------------------------------|-------------------------------------------------| +| "A helpful assistant" | Too vague; matches everything | +| "Does stuff with code" | No activation signal | +| "NOT for Python" | Negations don't help triggering, only exclusion | +| "General purpose tool" | Competes with every other contribution | +| Description > 500 chars | Loses signal in noise | + +### Negative Triggers + +If your contribution should NOT activate in certain contexts, add a "Do NOT use" line +in the body (not the frontmatter description): + +```markdown +Do NOT use when: +- The user is working with Python (use python-linter instead) +- The task is purely about UI styling with no logic +``` + +## Dos and Don'ts + +### Do + +- Write instructions in imperative mood ("Check the file", "Run the command") +- Be specific about file types, languages, and tools +- Include concrete examples of input/output in the body +- Test that your description uniquely identifies your contribution's purpose +- Keep the body under 2000 lines; link to external docs for exhaustive references +- Use markdown tables for structured mappings +- Provide a "Quick Start" section for complex agents + +### Don't + +- Include platform-specific tool schemas (JSON function definitions, MCP tool lists) +- Reference internal tool names (`Read`, `Edit`, `Bash`) - describe actions naturally +- Use XML tags in the body (``, ``, ``) +- Assume a specific model or context window size +- Include secrets, API keys, or environment-specific paths +- Write in first person ("I will...") - use imperative ("Check...", "Generate...") +- Duplicate content that belongs in a separate skill (link to it instead) + +## Concept Mapping: Claude/Codex to Copilot + +Use this table when migrating existing definitions to Copilot format. + +| Source Platform | Source Concept | Copilot Equivalent | +|-----------------|------------------------------------|-----------------------------------------------------| +| Claude | `SKILL.md` with behavior | `.agent.md` | +| Claude | `SKILL.md` with knowledge | `SKILL.md` (same name, different conventions) | +| Claude | `subagent_type` field | Separate `.agent.md` file per agent type | +| Claude | `isolation: "worktree"` | Body instruction: "Create a temporary branch" | +| Claude | `model: "opus"` / `"sonnet"` | Remove entirely; Copilot chooses model | +| Claude | `tools: [Bash, Read, Edit]` | Implicit; describe actions in prose | +| Claude | MCP server tool (`mcp__x__y`) | Body note: "Requires MCP server X" | +| Claude | `` XML blocks | Markdown code blocks with language tags | +| Claude | Skill `args` parameter | Body section: "Expected input" | +| Claude | `/skill-name` invocation | Automatic activation via description matching | +| Claude | `run_in_background` flag | Body instruction: "Can proceed asynchronously" | +| Codex | `functions:` array | Natural-language tool descriptions in body | +| Codex | `triggers:` keyword array | Folded into `description` field | +| Codex | `config.json` activation rules | `description` field + body "Do NOT use" section | +| Codex | OAuth/API plugin configuration | Not portable; note as prerequisite in body | +| Codex | JSON schema for parameters | Prose description of expected input | +| Codex | `shell_command` function type | Body instruction: "Run: `command`" | +| Both | Multi-agent orchestration | Multiple `.agent.md` files + linking in descriptions| +| Both | Conditional tool selection | Decision tree in body prose | + +## File Organization + +``` +awesome-copilot/ + agents/ + my-agent-name.agent.md # One file per agent + skills/ + my-skill-name/ + SKILL.md # One SKILL.md per skill directory +``` + +- Agent files live flat in `agents/` +- Skills get their own directory under `skills/` (allows supporting files) +- Names must match: directory name = `name` field in frontmatter +- Use kebab-case everywhere + +## Validation Checklist + +Before submitting a contribution: + +1. [ ] Frontmatter parses as valid YAML (test with `yq` or similar) +2. [ ] `name` is kebab-case and matches the filename/directory +3. [ ] `description` would let a machine decide when to activate (not just a human) +4. [ ] Body is self-contained - no "see original docs" dangling references +5. [ ] No platform-specific artifacts (XML tags, JSON schemas, tool arrays) +6. [ ] Instructions use imperative mood throughout +7. [ ] Examples use markdown code blocks, not custom markup +8. [ ] File is under 2000 lines +9. [ ] No secrets, absolute paths, or environment-specific values +10. [ ] Contribution does not duplicate an existing one in the repo From 340a51563106df241cdd9fc0f686a7fb069a2a5f Mon Sep 17 00:00:00 2001 From: Mammon Baloch <154027819+starlightretailceo@users.noreply.github.com> Date: Sat, 20 Jun 2026 19:08:52 -0700 Subject: [PATCH 2/2] chore: regenerate README index files Co-Authored-By: Claude Opus 4.6 (1M context) --- docs/README.agents.md | 1 + docs/README.skills.md | 1 + skills/copilot-skill-authoring/SKILL.md | 2 +- 3 files changed, 3 insertions(+), 1 deletion(-) diff --git a/docs/README.agents.md b/docs/README.agents.md index baf0e7d61..1220246ba 100644 --- a/docs/README.agents.md +++ b/docs/README.agents.md @@ -211,6 +211,7 @@ See [CONTRIBUTING.md](../CONTRIBUTING.md#adding-agents) for guidelines on how to | [Senior Cloud Architect](../agents/arch.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Farch.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Farch.agent.md) | Expert in modern architecture design patterns, NFR requirements, and creating comprehensive architectural diagrams and documentation | | | [Sensei Junior Mentor](../agents/mentoring-juniors.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fmentoring-juniors.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fmentoring-juniors.agent.md) | Socratic mentor for junior developers. Guides through questions, never gives direct answers. Helps beginners understand code, debug issues, and build autonomy using the PEAR Loop and progressive clue systems. | | | [Shopify Expert](../agents/shopify-expert.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fshopify-expert.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fshopify-expert.agent.md) | Expert Shopify development assistant specializing in theme development, Liquid templating, app development, and Shopify APIs | | +| [Skill Format Migrator](../agents/skill-format-migrator.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fskill-format-migrator.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fskill-format-migrator.agent.md) | Converts AI agent and skill definitions from Claude Agent Skills (SKILL.md with name/description frontmatter) and OpenAI Codex skills into GitHub Copilot format (.agent.md for personas with tools, SKILL.md for reusable knowledge). Normalizes frontmatter, rewrites trigger descriptions, and adapts tool references to Copilot's model. Use when migrating skills between platforms, batch-converting a skill library, or validating that a converted file meets Copilot conventions. | | | [Software Engineer Agent](../agents/software-engineer-agent-v1.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fsoftware-engineer-agent-v1.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fsoftware-engineer-agent-v1.agent.md) | Expert-level software engineering agent. Deliver production-ready, maintainable code. Execute systematically and specification-driven. Document comprehensively. Operate autonomously and adaptively. | | | [Specification](../agents/specification.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fspecification.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fspecification.agent.md) | Generate or update specification documents for new or existing functionality. | | | [Stackhawk Security Onboarding](../agents/stackhawk-security-onboarding.agent.md)
[![Install in VS Code](https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fstackhawk-security-onboarding.agent.md)
[![Install in VS Code Insiders](https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white)](https://aka.ms/awesome-copilot/install/agent?url=vscode-insiders%3Achat-agent%2Finstall%3Furl%3Dhttps%3A%2F%2Fraw.githubusercontent.com%2Fgithub%2Fawesome-copilot%2Fmain%2Fagents%2Fstackhawk-security-onboarding.agent.md) | Automatically set up StackHawk security testing for your repository with generated configuration and GitHub Actions workflow | stackhawk-mcp
[![Install MCP](https://img.shields.io/badge/Install-VS_Code-0098FF?style=flat-square)](https://aka.ms/awesome-copilot/install/mcp-vscode?name=stackhawk-mcp&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22stackhawk-mcp%22%5D%2C%22env%22%3A%7B%7D%7D)
[![Install MCP](https://img.shields.io/badge/Install-VS_Code_Insiders-24bfa5?style=flat-square)](https://aka.ms/awesome-copilot/install/mcp-vscodeinsiders?name=stackhawk-mcp&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22stackhawk-mcp%22%5D%2C%22env%22%3A%7B%7D%7D)
[![Install MCP](https://img.shields.io/badge/Install-Visual_Studio-C16FDE?style=flat-square)](https://aka.ms/awesome-copilot/install/mcp-visualstudio/mcp-install?%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22stackhawk-mcp%22%5D%2C%22env%22%3A%7B%7D%7D) | diff --git a/docs/README.skills.md b/docs/README.skills.md index a60d396d9..6392b75ad 100644 --- a/docs/README.skills.md +++ b/docs/README.skills.md @@ -103,6 +103,7 @@ See [CONTRIBUTING.md](../CONTRIBUTING.md#adding-skills) for guidelines on how to | [copilot-cli-quickstart](../skills/copilot-cli-quickstart/SKILL.md)
`gh skills install github/awesome-copilot copilot-cli-quickstart` | Use this skill when someone wants to learn GitHub Copilot CLI from scratch. Offers interactive step-by-step tutorials with separate Developer and Non-Developer tracks, plus on-demand Q&A. Just say "start tutorial" or ask a question! Note: This skill targets GitHub Copilot CLI specifically and uses CLI-specific tools (ask_user, sql, fetch_copilot_cli_documentation). | None | | [copilot-instructions-blueprint-generator](../skills/copilot-instructions-blueprint-generator/SKILL.md)
`gh skills install github/awesome-copilot copilot-instructions-blueprint-generator` | Technology-agnostic blueprint generator for creating comprehensive copilot-instructions.md files that guide GitHub Copilot to produce code consistent with project standards, architecture patterns, and exact technology versions by analyzing existing codebase patterns and avoiding assumptions. | None | | [copilot-sdk](../skills/copilot-sdk/SKILL.md)
`gh skills install github/awesome-copilot copilot-sdk` | Build agentic applications with GitHub Copilot SDK. Use when embedding AI agents in apps, creating custom tools, implementing streaming responses, managing sessions, connecting to MCP servers, or creating custom agents. Triggers on Copilot SDK, GitHub SDK, agentic app, embed Copilot, programmable agent, MCP server, custom agent. | None | +| [copilot-skill-authoring](../skills/copilot-skill-authoring/SKILL.md)
`gh skills install github/awesome-copilot copilot-skill-authoring` | Reference guide for authoring GitHub Copilot contributions in .agent.md and SKILL.md formats. Covers the frontmatter contract, format selection criteria, trigger-description patterns, dos and don'ts, and a mapping table from Claude/Codex concepts to Copilot equivalents. Use when creating new Copilot skills or agents, reviewing contributions for correctness, or migrating definitions from other AI platforms. | None | | [copilot-spaces](../skills/copilot-spaces/SKILL.md)
`gh skills install github/awesome-copilot copilot-spaces` | Use Copilot Spaces to provide project-specific context to conversations. Use this skill when users mention a "Copilot space", want to load context from a shared knowledge base, discover available spaces, or ask questions grounded in curated project documentation, code, and instructions. | None | | [copilot-usage-metrics](../skills/copilot-usage-metrics/SKILL.md)
`gh skills install github/awesome-copilot copilot-usage-metrics` | Retrieve and display GitHub Copilot usage metrics for organizations and enterprises using the GitHub CLI and REST API. | `get-enterprise-metrics.sh`
`get-enterprise-user-metrics.sh`
`get-org-metrics.sh`
`get-org-user-metrics.sh` | | [cosmosdb-datamodeling](../skills/cosmosdb-datamodeling/SKILL.md)
`gh skills install github/awesome-copilot cosmosdb-datamodeling` | Step-by-step guide for capturing key application requirements for NoSQL use-case and produce Azure Cosmos DB Data NoSQL Model design using best practices and common patterns, artifacts_produced: "cosmosdb_requirements.md" file and "cosmosdb_data_model.md" file | None | diff --git a/skills/copilot-skill-authoring/SKILL.md b/skills/copilot-skill-authoring/SKILL.md index d4f396ccf..18d74dd80 100644 --- a/skills/copilot-skill-authoring/SKILL.md +++ b/skills/copilot-skill-authoring/SKILL.md @@ -120,7 +120,7 @@ It must be precise, specific, and machine-parseable. | "Does stuff with code" | No activation signal | | "NOT for Python" | Negations don't help triggering, only exclusion | | "General purpose tool" | Competes with every other contribution | -| Description > 500 chars | Loses signal in noise | +| Description over 500 chars | Loses signal in noise | ### Negative Triggers