Skip to content

Add missing post-action verification to two skills#626

Open
milann wants to merge 1 commit intoobra:mainfrom
milann:fix/verifiability-gaps
Open

Add missing post-action verification to two skills#626
milann wants to merge 1 commit intoobra:mainfrom
milann:fix/verifiability-gaps

Conversation

@milann
Copy link

@milann milann commented Mar 6, 2026

While working through the workflow, I noticed two skills that complete irreversible or hard-to-reverse actions without confirming the outcome.

finishing-a-development-branch — Option 2

After git push and gh pr create, the skill moves straight to worktree cleanup with no confirmation that either operation succeeded. A failed push (e.g. rejected by remote) or a malformed gh pr create call would go unnoticed.

Added two verification steps after the push/create sequence:

  • git log origin/<branch> -1 --oneline — confirms the push landed
  • gh pr view --json url,title,state — confirms the PR was created

Also added to Red Flags: "Assume push/PR succeeded without checking" and to Always: "Confirm PR was created with gh pr view after Option 2".

dispatching-parallel-agents — Integration verification

The Verification section listed "Run full suite" as bullet point 3 of 4, at the same visual weight as "Spot check". Nothing indicated it was a blocking condition — an agent could read past it and report completion after just reviewing summaries.

Made the requirement unambiguous: added a bash block showing the command, and an explicit stop condition if the suite fails. Added "Never report integration complete without step 3 output in hand."

Both changes are minimal and follow the existing style of the skills.

…spatching-parallel-agents

finishing-a-development-branch Option 2 (Push + PR) had no confirmation
that the push and PR creation actually succeeded. Added git log and
gh pr view checks after the push/create sequence.

dispatching-parallel-agents listed 'run full suite' as a bullet point
in the Verification section but didn't mark it as a hard gate. Agents
could return with conflicting changes and the skill gave no clear signal
to stop if the suite failed. Made the requirement explicit with a
blocking condition.
@coderabbitai
Copy link

coderabbitai bot commented Mar 6, 2026

📝 Walkthrough

Walkthrough

Two skill documentation files have been updated with stricter verification protocols: one adds mandatory multi-step post-agent verification with explicit failure handling for parallel agent dispatching, while the other introduces push and PR creation confirmation steps for branch finishing workflows.

Changes

Cohort / File(s) Summary
Skills Verification Documentation
skills/dispatching-parallel-agents/SKILL.md, skills/finishing-a-development-branch/SKILL.md
Enhanced verification workflows with mandatory confirmation steps. Added stricter post-agent verification protocol with explicit failure gates and re-dispatch logic. Introduced push verification and PR creation confirmations with specific command examples for validation.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Suggested reviewers

  • yoavsion

Poem

🐰 Hoppy hops with glee,
Documentation now sees clearly!
Verify each step with care,
Confirm the push, the PR fair,
No leaps without proof so dear! 🐇✨

🚥 Pre-merge checks | ✅ 3
✅ Passed checks (3 passed)
Check name Status Explanation
Title check ✅ Passed The title 'Add missing post-action verification to two skills' accurately and clearly describes the main change in the changeset, capturing the core contribution.
Description check ✅ Passed The description is directly related to the changeset, providing detailed context about why verification steps were added to both skills and explaining the specific improvements made.
Docstring Coverage ✅ Passed No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✏️ Tip: You can configure your own custom pre-merge checks in the settings.

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Post copyable unit tests in a comment

Tip

Try Coding Plans. Let us write the prompt for your AI agent so you can ship faster (with fewer bugs).
Share your feedback on Discord.


Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands and usage tips.

Copy link

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

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

🧹 Nitpick comments (1)
skills/finishing-a-development-branch/SKILL.md (1)

95-96: Consider adding failure handling guidance for push verification.

The verification command git log origin/<feature-branch> -1 --oneline will confirm the push succeeded, but there's no explicit instruction on what to do if this command fails (e.g., branch not found on remote). While the command itself is correct, adding a brief note about stopping and investigating if the verification fails would make the workflow more robust.

📋 Suggested enhancement to add failure handling
 # Verify push succeeded
 git log origin/<feature-branch> -1 --oneline
+
+# If verification fails, stop and check push errors before proceeding
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.

In `@skills/finishing-a-development-branch/SKILL.md` around lines 95 - 96, Add
explicit failure-handling guidance for the verification step that uses `git log
origin/<feature-branch> -1 --oneline`: instruct the user that if the command
returns no result or errors (branch not found), they should abort further steps,
run `git fetch` and re-check remote refs, confirm the push command used (e.g.,
`git push origin HEAD:<feature-branch>`), and resolve remote branch or
permission issues before proceeding; include this note immediately after the
`git log origin/<feature-branch> -1 --oneline` line in SKILL.md so readers know
to stop and investigate on failure.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@skills/finishing-a-development-branch/SKILL.md`:
- Around line 95-96: Add explicit failure-handling guidance for the verification
step that uses `git log origin/<feature-branch> -1 --oneline`: instruct the user
that if the command returns no result or errors (branch not found), they should
abort further steps, run `git fetch` and re-check remote refs, confirm the push
command used (e.g., `git push origin HEAD:<feature-branch>`), and resolve remote
branch or permission issues before proceeding; include this note immediately
after the `git log origin/<feature-branch> -1 --oneline` line in SKILL.md so
readers know to stop and investigate on failure.

ℹ️ Review info
⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: f20dc9ff-143b-43d0-b1c2-f4ae7672a74e

📥 Commits

Reviewing files that changed from the base of the PR and between e4a2375 and 2a56f08.

📒 Files selected for processing (2)
  • skills/dispatching-parallel-agents/SKILL.md
  • skills/finishing-a-development-branch/SKILL.md

IgorTavcar added a commit to IgorTavcar/superpowers that referenced this pull request Mar 9, 2026
…hing (obra#626)

Adds test-suite gate for parallel agent completion and explicit push/PR
confirmation commands for branch finishing.

Upstream PR: obra#626

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
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.

1 participant