Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 6 additions & 3 deletions skills/github-agentic-workflows/references/authoring.md
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,8 @@ Useful frontmatter fields for professional workflows:
7. `mcp:` when the workflow needs custom tools or integrations via Model Context Protocol servers.
8. `max-ai-credits:` to cap the AI Credits (AIC) a single run may consume and prevent runaway inference spend.

When selecting `toolsets:`, use the GitHub tools reference at `https://github.github.com/gh-aw/reference/github-tools/` to understand which toolsets expose which operations before committing to a minimal set.

Practical default: use `engine: copilot` unless the repository explicitly needs another model provider and is already wired for that provider's secret and runtime expectations.

## Authoring Style
Expand Down Expand Up @@ -126,6 +128,7 @@ For authoring with agents:

1. Use `gh aw init` to prepare the repository. For agent-assisted setup, use `https://raw.githubusercontent.com/github/gh-aw/main/install.md`.
2. Use the debugging prompt at `https://raw.githubusercontent.com/github/gh-aw/main/debug.md` for failed runs.
3. Use the planner or dictation prompts from the official docs when task shaping is the bottleneck rather than workflow syntax.
4. Use the interactive workshop at `https://github.github.com/gh-aw/workshop/` for hands-on exercises that cover installation, authoring, and running agentic workflows end to end.
5. If strict mode rejects external-source domains that are still required for the task, add deterministic pre-steps that fetch those sources to local files and instruct the agent to use the local copies.
3. Use the official GH-AW agent file at `https://raw.githubusercontent.com/github/gh-aw/main/.github/aw/github-agentic-workflows.md` when the repository itself uses a coding agent for workflow authoring and needs the canonical authoring guidance in a form the agent can read directly.
4. Use the planner or dictation prompts from the official docs when task shaping is the bottleneck rather than workflow syntax.
5. Use the interactive workshop at `https://github.github.com/gh-aw/workshop/` for hands-on exercises that cover installation, authoring, and running agentic workflows end to end.
6. If strict mode rejects external-source domains that are still required for the task, add deterministic pre-steps that fetch those sources to local files and instruct the agent to use the local copies.
10 changes: 5 additions & 5 deletions skills/github-agentic-workflows/references/examples.md
Original file line number Diff line number Diff line change
Expand Up @@ -129,12 +129,12 @@ Review existing workflows against these questions:

High-value workflow families from the GH-AW docs and gallery include:

1. IssueOps and LabelOps.
1. IssueOps and LabelOps — see `https://github.github.com/gh-aw/guides/ai-issue-triage/`.
2. DailyOps and Monitoring.
3. Continuous Documentation — documentation maintenance and consistency.
3. Continuous Documentation — documentation maintenance and consistency. See `https://github.github.com/gh-aw/guides/docs-automation/`.
4. Continuous Improvement — daily code simplification, refactoring, and style improvements.
5. Quality and testing hygiene.
6. Multi-repository coordination.
7. ProjectOps and status reporting.
5. Quality and testing hygiene — see `https://github.github.com/gh-aw/guides/automated-pr-review/`.
6. Multi-repository coordination — see `https://github.github.com/gh-aw/examples/multi-repo/`.
7. ProjectOps and status reporting — see `https://github.github.com/gh-aw/guides/ai-release-notes/`.
8. Orchestration and task delegation.
9. Assign-to-Copilot or create-agent-session follow-up automation.