Skip to content

docs: add agent skill for running workflows locally with act - #6155

Open
gdevenyi wants to merge 3 commits into
nektos:masterfrom
gdevenyi:add-act-skill
Open

docs: add agent skill for running workflows locally with act#6155
gdevenyi wants to merge 3 commits into
nektos:masterfrom
gdevenyi:add-act-skill

Conversation

@gdevenyi

@gdevenyi gdevenyi commented Aug 6, 2026

Copy link
Copy Markdown

Summary

Adds an agent skill at skills/act/SKILL.md so coding agents (Claude Code and compatible tools) know how to use act to run and debug GitHub Actions workflows locally. The skills/ location lets users install it into any project with the skills CLI: npx skills add nektos/act.

The skill deliberately stays small: act --help remains the authoritative flag reference, and the file only carries the run workflow (scout with -l/-n, narrow with -j/-W/--matrix, supply secrets/inputs/event payloads) plus the gotchas --help does not cover — runner image gaps vs GitHub's hosted images, the first-run image prompt in non-interactive sessions, .actrc, ACT=true detection, the artifact server, and --bind writing into the host checkout.

Motivation

Agents asked to "reproduce this CI failure locally" routinely stumble on the same act-specific traps (missing tools in the medium image, no artifact server, the interactive first-run prompt). Shipping the skill in-repo gives every act user's agent the same tribal knowledge.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Dye4h9PhFEjURQvp6CyQXy

Adds a Claude Code / agent skill at .claude/skills/act/SKILL.md, alongside
the existing CLAUDE.md, so coding agents working in any checkout that
installs it know how to scout, run, and debug workflows with act - and
the gotchas that act --help does not cover (runner image gaps, first-run
image prompt, .actrc, ACT=true detection, artifact server).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Dye4h9PhFEjURQvp6CyQXy
Copilot AI lite review requested due to automatic review settings August 6, 2026 20:05
gdevenyi and others added 2 commits August 6, 2026 16:06
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Dye4h9PhFEjURQvp6CyQXy
@pull-request-size pull-request-size Bot added size/M and removed size/S labels Aug 6, 2026

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR adds a new Claude/agent skill document to help coding agents use act to run and debug GitHub Actions workflows locally, capturing common workflow steps and local-run “gotchas” that aren’t obvious from act --help.

Changes:

  • Add .claude/skills/act/SKILL.md with a concise “how to run” workflow for act (scouting, narrowing, secrets/inputs, verbosity).
  • Document common local pitfalls (runner image differences, first-run image prompt, .actrc, ACT=true, artifact/cache servers, and --bind behavior).

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants