Skip to content

[Copilot Dreams] Diagnose slow runs and locate actual CI checks #270

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Copilot wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[Copilot Dreams] Diagnose slow runs and locate actual CI checks #270

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
2 changes: 1 addition & 1 deletion AGENTS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,7 +46,7 @@ claim in both.
here; it can't be expressed in the catalog.
- **Go command test** (`cmd/**/*_test.go` with `httpmock`) — full CLI wiring
end to end, but with a stubbed resolver so it stays deterministic. Runs in
the blocking `test` CI job with no token. Reach for this to prove the
the blocking `go-test` CI job with no token. Reach for this to prove the

Copilot AI Nov 5, 2026

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

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

🛠️ Find the Go check: The old job name could send Copilot looking for a check identifier that is not defined. The workflow names the job go-test; the command tests still run there without a supplied API token.

command plumbs its pieces together.
- **Catalog scenario** (`test/scenarios/catalog.yml`, run by the Ruby
harness) — the *real compiled binary* against the real API or a stub
Expand Down
7 changes: 5 additions & 2 deletions skills/gh-actions-lock/SKILL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -135,8 +135,8 @@ category name alone.

Copy [`.github/workflows/verify-lockfile.yml`](https://github.com/github/gh-actions-lock/blob/main/.github/workflows/verify-lockfile.yml)
into the repository, then run `gh actions-lock` once so the new workflow is locked too. It runs
`gh actions-lock --verify` on pull requests, pushes to `main`, and a weekly schedule that catches
re-pointed tags. The job's `GITHUB_TOKEN` reads public actions and the repository itself; actions
`gh actions-lock --verify` on pull requests and pushes to `main`. The job's `GITHUB_TOKEN` reads

Copilot AI Nov 5, 2026

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

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

🛠️ Avoid assuming weekly checks: The old claim could make Copilot rely on verification after repository activity stops. The copied workflow triggers only on pull requests and pushes to main; those checks remain, but the template supplies no weekly run.

public actions and the repository itself; actions
from other private or internal repositories need a token that can read them.

## Limits
Expand All @@ -149,6 +149,9 @@ from other private or internal repositories need a token that can read them.

## Learn more

To diagnose a slow online run, add `--profile <dir>` and read the completed run's

Copilot AI Nov 5, 2026

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

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

🛠️ Use recorded performance data: A help pointer alone leaves Copilot to derive timing and request counts from debug logs. For slow online runs, the existing profiler writes phase timings on completion and instrumented request counts and latency, avoiding manual log counting without replacing the reproduction.

stderr summary for phase timings, HTTP request counts, and latency.

`gh actions-lock --help` is authoritative for flags. Failed runs print the path of a saved run
log; attach it to bug reports. Start from these, plus each finding's `doc_url`:

Expand Down
Loading