Skip to content

Comments

Bug Hunter: require agent-authored reproduction, prohibit running existing test suite#189

Merged
strawgate merged 2 commits intomainfrom
copilot/fix-bug-hunter-logic
Feb 20, 2026
Merged

Bug Hunter: require agent-authored reproduction, prohibit running existing test suite#189
strawgate merged 2 commits intomainfrom
copilot/fix-bug-hunter-logic

Conversation

Copy link
Contributor

Copilot AI commented Feb 20, 2026

Bug Hunter was running the existing test suite (make test, pytest, etc.) and reporting failures as bugs. The intent is to find logic bugs through code analysis and prove them with a new minimal failing test — not to be a flaky-test detector.

Prompt changes (gh-aw-bug-hunter.md)

  • Reproduce step: must write a new script/test targeting the specific bug found via code analysis; explicitly bans running the existing test suite ("if you did not write the test, a failure is not your finding")
  • What to Look For: adds logic errors (wrong conditionals, off-by-one, wrong variable, missing edge-case handling) as the primary target class
  • What to Skip: adds — "Test suite failures from running existing tests — do not run make test or similar and report the output as a bug"
  • Quality Gate: adds noop trigger — "Your only evidence is an existing test failure you did not cause"
  • Issue format: renames Suggested Failing TestFailing Test; requires full code of the new test, not just an outline

Lock file regenerated via make compile.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

  • https://api.github.com/repos/ruby/setup-ruby/git/ref/tags/v1
    • Triggering command: /usr/bin/gh gh api /repos/ruby/setup-ruby/git/ref/tags/v1 --jq .object.sha -c=4 -nolocalimports -importcfg /tmp/go-build3071520046/b334/importcfg -pack /home/REDACTED/go/pkg/mod/github.com/spf13/pflag@v1.0.10/bool.go /home/REDACTED/go/pkg/mod/github.com/spf13/pflag@v1.0.10/bool_func.go ux-a�� 0.1-go1.25.7.lin--gdwarf-5 (http block)
    • Triggering command: /usr/bin/gh gh api /repos/ruby/setup-ruby/git/ref/tags/v1 --jq .object.sha /tmp/go-build3071520046/b338/symabis -c=4 -nolocalimports -importcfg /tmp/go-build3071520046/b338/importcfg -pack -asmhdr (http block)
    • Triggering command: /usr/bin/gh gh api /repos/ruby/setup-ruby/git/ref/tags/v1 --jq .object.sha /tmp/go-build3071520046/b337/symabis -c=4 -nolocalimports -importcfg /tmp/go-build3071520046/b337/importcfg -pack -asmhdr (http block)
  • https://api.github.com/user
    • Triggering command: /usr/bin/gh gh api user --jq .login go1.25.7 -c=4 -nolocalimports -importcfg /tmp/go-build3071520046/b325/importcfg -pack /home/REDACTED/go/pkg/mod/golang.org/toolchain@v0.0.1-go1.25.7.linux-amd64/src/testing/allocs.go (http block)

If you need me to access, download, or install something from one of these locations, you can either:

Original prompt

This section details on the original issue you should resolve

<issue_title>Bug Hunter Improvements</issue_title>
<issue_description>See https://github.com/elastic/beats/actions/runs/22206903339

We already know tests are passing, so why is it running tests?

The goal is to identify a bug / logic issue / etc</issue_description>

Comments on the Issue (you are @copilot in this section)

@strawgate I want it to find a bug, build a test that shows that the bug exists, etc. It's not supposed to be just running tests hoping one fails and then calling that a bug

Custom agent used: default-agent
The default agent


💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

Copilot AI changed the title [WIP] Fix bug hunter improvements for test execution Bug Hunter: require agent-authored reproduction, prohibit running existing test suite Feb 20, 2026
Copilot AI requested a review from strawgate February 20, 2026 01:42
@strawgate strawgate force-pushed the copilot/fix-bug-hunter-logic branch from 48872c1 to 341171f Compare February 20, 2026 01:48
@strawgate strawgate marked this pull request as ready for review February 20, 2026 01:48
Copy link
Contributor

@github-actions github-actions bot left a comment

Choose a reason for hiding this comment

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

Verdict: APPROVE. I found no actionable issues in this PR.


What is this? | From workflow: PR Review Agent

Give us feedback! React with 🚀 if perfect, 👍 if helpful, 👎 if not. for issue #189

@github-actions
Copy link
Contributor

The failure is from Verify compiled workflowsEnsure lock files are up to date; make compile rewrites generated workflow artifacts, so the branch is out of sync with generated outputs.

Root cause observed in the failed log and reproduced locally with make compile:

  • many tracked files under .github/workflows/*.lock.yml and .github/workflows/agentics-maintenance.yml become modified
  • two generated trigger files are missing from git: .github/workflows/trigger-flaky-test-triage.yml and .github/workflows/trigger-issue-triage-pr.yml

Validation run:

  • make compile (reproduced the same drift)

Recommended fix on the PR branch:

  1. Run make compile
  2. Commit all generated workflow changes (including the two missing trigger files above)
  3. Push and re-run CI

What is this? | From workflow: PR Checks Fixer

Give us feedback! React with 🚀 if perfect, 👍 if helpful, 👎 if not.

@github-actions
Copy link
Contributor

CI failure is from Verify compiled workflowsEnsure lock files are up to date, not from tests.

make compile succeeds, but the step immediately fails because the repo is dirty after compile:

  • Working tree dirty after make compile.
  • Modified: many gh-aw-*.lock.yml files plus .github/workflows/agentics-maintenance.yml
  • Untracked: .github/workflows/trigger-flaky-test-triage.yml and .github/workflows/trigger-issue-triage-pr.yml

This means generated workflow artifacts are out of sync with the source templates in this branch.

Recommended fix (minimal)

  1. Rebase/merge latest main into copilot/fix-bug-hunter-logic.
  2. Run make compile locally.
  3. Commit all resulting generated workflow updates (including newly generated trigger files).
  4. Push and re-run CI.

What I ran

  • Inspected workflow run 22207976958 jobs and failed job logs.
  • No additional local test suite was run (failure occurs before tests; this is a generation-sync check).

What is this? | From workflow: PR CI Detective

Give us feedback! React with 🚀 if perfect, 👍 if helpful, 👎 if not.

@strawgate strawgate merged commit 8efe885 into main Feb 20, 2026
10 of 11 checks passed
@strawgate strawgate deleted the copilot/fix-bug-hunter-logic branch February 20, 2026 02:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Bug Hunter Improvements

2 participants