You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: AGENTS.md
+10-30Lines changed: 10 additions & 30 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -2,10 +2,9 @@
2
2
3
3
## Project Overview
4
4
5
-
The Awesome GitHub Copilot repository is a community-driven collection of custom agents, prompts, and instructions designed to enhance GitHub Copilot experiences across various domains, languages, and use cases. The project includes:
5
+
The Awesome GitHub Copilot repository is a community-driven collection of custom agents and instructions designed to enhance GitHub Copilot experiences across various domains, languages, and use cases. The project includes:
6
6
7
7
-**Agents** - Specialized GitHub Copilot agents that integrate with MCP servers
8
-
-**Prompts** - Task-specific prompts for code generation and problem-solving
9
8
-**Instructions** - Coding standards and best practices applied to specific file patterns
10
9
-**Skills** - Self-contained folders with instructions and bundled resources for specialized tasks
11
10
-**Hooks** - Automated workflows triggered by specific events during development
@@ -17,7 +16,6 @@ The Awesome GitHub Copilot repository is a community-driven collection of custom
@@ -55,23 +53,16 @@ npm run skill:create -- --name <skill-name>
55
53
56
54
## Development Workflow
57
55
58
-
### Working with Agents, Prompts, Instructions, Skills, and Hooks
56
+
### Working with Agents, Instructions, Skills, and Hooks
59
57
60
-
All agent files (`*.agent.md`), prompt files (`*.prompt.md`), and instruction files (`*.instructions.md`) must include proper markdown front matter. Agent Skills are folders containing a `SKILL.md` file with frontmatter and optional bundled assets. Hooks are folders containing a `README.md` with frontmatter and a `hooks.json` configuration file:
58
+
All agent files (`*.agent.md`) and instruction files (`*.instructions.md`) must include proper markdown front matter. Agent Skills are folders containing a `SKILL.md` file with frontmatter and optional bundled assets. Hooks are folders containing a `README.md` with frontmatter and a `hooks.json` configuration file:
61
59
62
60
#### Agent Files (*.agent.md)
63
61
- Must have `description` field (wrapped in single quotes)
64
62
- File names should be lower case with words separated by hyphens
65
63
- Recommended to include `tools` field
66
64
- Strongly recommended to specify `model` field
67
65
68
-
#### Prompt Files (*.prompt.md)
69
-
- Must have `agent` field (value should be `'agent'` wrapped in single quotes)
70
-
- Must have `description` field (wrapped in single quotes, not empty)
71
-
- File names should be lower case with words separated by hyphens
72
-
- Recommended to specify `tools` if applicable
73
-
- Strongly recommended to specify `model` field
74
-
75
66
#### Instruction Files (*.instructions.md)
76
67
- Must have `description` field (wrapped in single quotes, not empty)
77
68
- Must have `applyTo` field specifying file patterns (e.g., `'**.js, **.ts'`)
@@ -102,12 +93,10 @@ All agent files (`*.agent.md`), prompt files (`*.prompt.md`), and instruction fi
102
93
- Each workflow is a standalone `.md` file in the `workflows/` directory
103
94
- Must have `name` field (human-readable name)
104
95
- Must have `description` field (wrapped in single quotes, not empty)
105
-
- Should have `triggers` field (array of trigger types, e.g., `['schedule', 'issues']`)
106
96
- Contains agentic workflow frontmatter (`on`, `permissions`, `safe-outputs`) and natural language instructions
107
97
- File names should be lower case with words separated by hyphens
108
98
- Only `.md` files are accepted — `.yml`, `.yaml`, and `.lock.yml` files are blocked by CI
109
-
- Optionally includes `tags` field for categorization
110
-
- Follow the [GitHub Agentic Workflows specification](https://github.github.com/gh-aw)
99
+
- Follow the [GitHub Agentic Workflows specification](https://github.github.com/gh-aw/reference/workflow-structure/)
111
100
112
101
#### Plugin Folders (plugins/*)
113
102
- Each plugin is a folder containing a `.github/plugin/plugin.json` file with metadata
@@ -120,9 +109,9 @@ All agent files (`*.agent.md`), prompt files (`*.prompt.md`), and instruction fi
120
109
121
110
### Adding New Resources
122
111
123
-
When adding a new agent, prompt, instruction, skill, hook, workflow, or plugin:
112
+
When adding a new agent, instruction, skill, hook, workflow, or plugin:
124
113
125
-
**For Agents, Prompts, and Instructions:**
114
+
**For Agents and Instructions:**
126
115
1. Create the file with proper front matter
127
116
2. Add the file to the appropriate directory
128
117
3. Update the README.md by running: `npm run build`
@@ -140,7 +129,7 @@ When adding a new agent, prompt, instruction, skill, hook, workflow, or plugin:
140
129
141
130
**For Workflows:**
142
131
1. Create a new `.md` file in `workflows/` with a descriptive name (e.g., `daily-issues-report.md`)
143
-
2. Include frontmatter with `name`, `description`, `triggers`, plus agentic workflow fields (`on`, `permissions`, `safe-outputs`)
132
+
2. Include frontmatter with `name` and `description`, plus agentic workflow fields (`on`, `permissions`, `safe-outputs`)
144
133
3. Compile with `gh aw compile --validate` to verify it's valid
145
134
4. Update the README.md by running: `npm run build`
146
135
5. Verify the workflow appears in the generated README
@@ -207,7 +196,7 @@ When creating a pull request:
207
196
3.**File naming**: Verify all new files follow the lower-case-with-hyphens naming convention
208
197
4.**Build check**: Run `npm run build` before committing to verify README generation
209
198
5.**Line endings**: **Always run `bash scripts/fix-line-endings.sh`** to normalize line endings to LF (Unix-style)
210
-
6.**Description**: Provide a clear description of what your agent/prompt/instruction does
199
+
6.**Description**: Provide a clear description of what your agent/instruction does
211
200
7.**Testing**: If adding a plugin, run `npm run plugin:validate` to ensure validity
212
201
213
202
### Pre-commit Checklist
@@ -222,13 +211,6 @@ Before submitting your PR, ensure you have:
222
211
223
212
### Code Review Checklist
224
213
225
-
For prompt files (*.prompt.md):
226
-
-[ ] Has markdown front matter
227
-
-[ ] Has `agent` field (value should be `'agent'` wrapped in single quotes)
228
-
-[ ] Has non-empty `description` field wrapped in single quotes
229
-
-[ ] File name is lower case with hyphens
230
-
-[ ] Includes `model` field (strongly recommended)
231
-
232
214
For instruction files (*.instructions.md):
233
215
-[ ] Has markdown front matter
234
216
-[ ] Has non-empty `description` field wrapped in single quotes
@@ -266,13 +248,11 @@ For workflow files (workflows/*.md):
266
248
-[ ] File has markdown front matter
267
249
-[ ] Has `name` field with human-readable name
268
250
-[ ] Has non-empty `description` field wrapped in single quotes
269
-
-[ ] Has `triggers` array field listing workflow trigger types
270
251
-[ ] File name is lower case with hyphens
271
252
-[ ] Contains `on` and `permissions` in frontmatter
272
253
-[ ] Workflow uses least-privilege permissions and safe outputs
273
254
-[ ] No `.yml`, `.yaml`, or `.lock.yml` files included
-[ ] Directory contains a `.github/plugin/plugin.json` file
@@ -295,7 +275,7 @@ This is a community-driven project. Contributions are welcome! Please see:
295
275
296
276
## MCP Server
297
277
298
-
The repository includes an MCP (Model Context Protocol) Server that provides prompts for searching and installing resources directly from this repository. Docker is required to run the server.
278
+
The repository includes an MCP (Model Context Protocol) Server for searching and installing resources directly from this repository. Docker is required to run the server.
0 commit comments