Prepares act-testable GitHub Actions workflow scenarios for this monorepo.
SkillScore91Maintained
Automatic
◆DevOps & SRE pack
Last verified 8/10/2026
What it changes
Passed 17 of 19 test cases with the skill, and 5 of 19 without it.
Without skill27%
With skill91%
Checks it passed that the model alone missed
Includes contract key act-scenarios: not required
Classifies the markdown validation workflow as Tier 1
Rejects running act directly against production workflow files
and 18 more in the full report
15 fixed5 passed either way1 still failing1 broke
Pass-rate lift
+64 pp
Tokens
+65%
Agent turns
0%
Response time
-46%
Model gemini-3.6-flashMeasured 8/10/2026v120 of 22 cases passed with skill · 19 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 14
Prompt
I want to trigger a manual dispatch test on my harness `.github/workflows/test-deploy.yml` with event payload `.claude/skills/act-testing/references/event-payloads/dispatch.json`. What is the canonical argument ordering for running `act`?
What was checked, without and with the skill
Checked
Without skill
With skill
Places the event JSON flag -e before the workflow flag -W in the act command
Missed
Passed
Appends --secret-file before --reuse at the end of the argument string
Missed
Passed
case-11 · gemini-3.6-flash · 8/10/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md169 lines6 filesincludes scripts
Act Testing Skill
Use this skill when a workflow change needs local validation with act.
Mandatory Validation Checkpoint
After creating or editing a workflow:
Assess if the workflow is Tier 1, Tier 2, or Tier 3 (see references/workflow-tiers.md).
Re-assess tier after every major scope change during iteration (e.g., removed jobs/branches/options).
If Tier 1 or Tier 2, create a temporary test-<workflow>.yml harness and run drift guard:
Assess whether act execution is possible in the current environment.
Provide the user with concrete test cases and suggested act commands to run manually.
If Tier 3, explicitly document why act testing is not applicable.
Never silently skip the testability decision.
Never run act directly against production *.yml files — always use a test-*.yml harness.
The agent prepares harnesses and validation guidance; the user runs act.
Tier Decision Rubric (use before preparing act guidance)
Use this quick rubric per changed workflow:
Tier 1 (fully testable): core behavior is internal control flow, expressions, and bash logic that can be validated locally.
Tier 2 (partially testable): workflow combines meaningful internal logic with external integrations; test internal logic with harness + mocks.
Tier 3 (not act-testable): workflow value is predominantly external orchestration/side effects (e.g., Vault-dependent auth, remote API mutation, deployment state), and local mocks would provide low confidence.
Decision rule:
If internal logic is a primary risk, prepare harness + scenario matrix (Tier 1/2).
If external orchestration is the primary risk and internal logic is minimal, Tier 3 rationale is acceptable.
For multi-workflow changes, classify each workflow explicitly; do not infer one workflow's tier from another.
Mocked-Signal Gate (mandatory)
Before committing to Tier 1/2 harnessing, estimate how much critical behavior is mocked in the harness:
If >=70% of critical path steps are mocked, do not default to Tier 1/2.
Diagnose failing GitHub Actions runs and propose fixes. Use when given a GHA run/job URL, or asked why CI failed, why a job failed, what broke in a run, or to fix a red check on a PR.