WIP - Fix cursor E2E flaky tests

[alishakawaguchi]

No description provided.

[cursor]

PR Summary

Low Risk
Mostly adds developer tooling/docs plus a small, low-impact E2E bootstrap change (directory creation) for Cursor CLI; no production runtime paths are modified.

Overview
Introduces a new local Claude plugin/skill suite for E2E work: /e2e orchestration plus phase commands (/e2e:triage-ci, /e2e:debug, /e2e:implement) with new procedures for CI triage (including re-run based classification) and fix verification.

Adds scripts/download-e2e-artifacts.sh to fetch and normalize GitHub Actions E2E artifacts for local analysis, and updates E2E/agent-integration docs to point at the new /e2e:debug workflow and CI triage flow.

Hardens the cursor-cli E2E agent bootstrap by pre-creating ~/.config/cursor to avoid ENOENT failures when Cursor writes config during trust acceptance; includes a tiny test data formatting tweak in common_test.go.

^{Written by Cursor Bugbot for commit 3e27584. Configure here.}

[Copilot]

Pull request overview

Adds tooling and documentation to improve debugging/triage of flaky E2E runs (especially Cursor), including a helper script for downloading CI artifacts and new .claude skill/plugin docs for a triage→fix workflow.

Changes:

• Add scripts/download-e2e-artifacts.sh to fetch and normalize GitHub Actions E2E artifacts locally.
• Update Cursor E2E agent bootstrap to pre-create the Cursor config directory to avoid runtime ENOENT failures.
• Add/update E2E triage/debug/implement skill + plugin documentation and refresh E2E README guidance.

Reviewed changes

Copilot reviewed 15 out of 16 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
scripts/download-e2e-artifacts.sh	New helper script to locate a run (latest/ID/URL), download artifacts, flatten wrapper dirs, and write .run-info.json.
e2e/agents/cursor_cli.go	Create Cursor config directory during bootstrap to reduce flaky failures.
e2e/README.md	Document local artifact download + new triage workflow references.
cmd/entire/cli/strategy/common_test.go	Minor formatting/alignment in tests.
.claude/skills/e2e/triage-ci.md	New CI triage procedure doc (download artifacts or rerun locally, classify flaky vs real-bug).
.claude/skills/e2e/implement.md	New procedure doc for applying fixes and verifying via scoped E2E runs.
.claude/skills/e2e/debug.md	Update/normalize debug procedure doc formatting.
.claude/skills/e2e/SKILL.md	Add orchestrator skill definition for the E2E triage→implement pipeline.
.claude/skills/agent-integration/*.md	Update references to use /e2e:debug instead of the old command name.
.claude/plugins/e2e/**	Add local plugin command wrappers + plugin metadata and README.

[Copilot]

With set -euo pipefail, this URL parsing pipeline will cause the script to exit immediately if the URL doesn’t contain /runs/<id> (grep returns non-zero), so the subsequent "Could not extract" check never runs. Capture the pipeline’s failure explicitly (e.g., via if ! run_id=...; then die ...; fi or || true) so invalid URLs produce the intended error message instead of an abrupt exit.

Suggested change

	run_id=$(echo "$input" | grep -oE '/runs/[0-9]+' | grep -oE '[0-9]+')
	if ! run_id=$(echo "$input" | grep -oE '/runs/[0-9]+' | grep -oE '[0-9]+'); then
	die "Could not extract run ID from URL: $input"
	fi

[Copilot]

The redirection 2>&1 >&2 doesn’t reliably enforce the contract that only the final absolute path is written to stdout; depending on evaluation order it can still leak output to stdout. Redirect the command’s stdout to stderr explicitly (and keep stderr on stderr) so callers can safely capture stdout.

Suggested change

	gh run download "$run_id" --dir "$dest" 2>&1 >&2 || die "Failed to download artifacts. They may have expired (retention: 7 days)."
	gh run download "$run_id" --dir "$dest" 1>&2 || die "Failed to download artifacts. They may have expired (retention: 7 days)."

[Copilot]

agents_found is computed via ls -d */ | ... under set -euo pipefail. If the download directory contains no subdirectories (e.g., no artifacts were present or the layout changes), the */ glob/ls will fail and the script will exit before writing .run-info.json. Consider using a glob with nullglob, find, or otherwise handling the empty case so the script fails with a clear message (or records an empty agent list) instead of exiting on ls.

Suggested change

	agents_found=$(cd "$dest" && ls -d */ 2>/dev/null | tr -d '/' | tr '\n' ', ' | sed 's/,$//')
	agents_found=$(cd "$dest" && find . -mindepth 1 -maxdepth 1 -type d -printf '%f\n' | tr '\n' ', ' | sed 's/, $//')

[Copilot]

This hard-codes the Linux-style config path ~/.config/cursor. To keep local E2E runs working across OSes (and respect XDG_CONFIG_HOME), prefer deriving the base config dir with os.UserConfigDir() and then appending cursor.

Suggested change

	home, err := os.UserHomeDir()
	if err != nil {
	return fmt.Errorf("get home dir: %w", err)
	}
	dir := filepath.Join(home, ".config", "cursor")
	cfgDir, err := os.UserConfigDir()
	if err != nil {
	return fmt.Errorf("get user config dir: %w", err)
	}
	dir := filepath.Join(cfgDir, "cursor")

[Copilot]

This references an e2e-triage skill at .claude/skills/e2e-triage/ and a /e2e-triage command, but this PR adds the E2E skills under .claude/skills/e2e/ and the plugin commands as /e2e:triage-ci (or the orchestrator /e2e). Update the paths/command names here so the README matches what actually exists in the repo.

Suggested change

	Use the `debug-e2e` skill (`.claude/skills/debug-e2e/`) for a structured workflow when investigating failures.
	Use the `e2e-triage` skill (`.claude/skills/e2e-triage/`) to automate full triage: download CI artifacts, classify failures as flaky vs real bug, and create PRs or GitHub issues. Run locally with `/e2e-triage` or see the automated CI workflow below.
	Use the E2E debug workflow in the `e2e` skill (`.claude/skills/e2e/`) for a structured workflow when investigating failures.
	Use the CI triage workflow in the same `e2e` skill to automate full triage: download CI artifacts, classify failures as flaky vs real bug, and create PRs or GitHub issues. Run locally with the `/e2e:triage-ci` plugin command (or invoke the orchestrator with `/e2e`), or see the automated CI workflow below.

[Copilot]

The README lists .github/workflows/e2e-triage.yml, but there’s no such workflow file in .github/workflows/ in this branch. Either add the workflow in this PR or remove/adjust this bullet so the documentation doesn’t point at a non-existent file.

Suggested change

- **`.github/workflows/e2e-triage.yml`** — Auto-triggers on E2E failure via `workflow_run`. Runs Claude Code (Opus) to download artifacts, classify failures, and create PRs (flaky) or issues (real bugs).

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

Comment @cursor review or bugbot run to trigger another review on this PR

[cursor]

Wrapper directory not removed after copy operation

Low Severity

In the restructuring loop, when the $agent directory already exists, cp -r copies from the e2e-artifacts-* wrapper into it but never removes the wrapper directory. The stale e2e-artifacts-* directory then gets picked up by the ls -d */ on line 92, polluting agents_found in .run-info.json with entries like e2e-artifacts-claude-code alongside claude-code.