feat(strands-command): add bug verifier agent#71
Open
JackYPCOnline wants to merge 1 commit into
Open
Conversation
Adds a new agent that triages bug reports by inspecting the implicated source, attempting a reproduction (the reporter's steps first, a derived repro as fallback), scoring urgency, and applying triage labels. Comments on the issue only when (1) the reporter's steps could not be validated but the agent produced its own working repro (shares it), or (2) neither path reproduced the bug (requests more information). - New SOP: task-bug-verifier.sop.md - Wire 'bug-verify' command through process-input.cjs - New additive write tool add_issue_labels in github_tools.py - Register the tool in agent_runner.py and write_executor.py Companion auto-trigger workflow and local-preview skill ship in the sdk-python repo as a separate PR.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation
Bug reports vary in quality. A maintainer typically has to (1) decide whether the
described behavior is plausible from the code, (2) try to reproduce it (often
reconstructing a runnable repro from prose), and (3) judge urgency before
prioritizing. This duplicates work across every report and delays response.
This PR adds an agent that performs that triage automatically: inspects the
implicated source, attempts the reporter's reproduction in the sandbox and
falls back to producing one of its own, scores urgency, and applies triage
labels. It comments on the issue in two cases only — to share a derived
reproduction the reporter didn't supply, or to ask for more information when
reproduction failed entirely.
Public API changes
A new
/strands bug-verifycommand (orbug verify) backed bytask-bug-verifier.sop.md, plus one new write tool:Additive (preserves existing labels), routes through the same deferred-write
path as the other write tools.
Verdict → action
bug-validated+P0–P3bug-validated+P0–P3bug-needs-infobug-cannot-reproduce+autoclose in 7 daysFollow-up
A companion auto-trigger workflow (on
[BUG]issue creation) and alocal-preview skill ship in
sdk-pythonas a separate PR, since they'rerepo-specific. Consumer repos need to create the new triage labels in repo
settings.