OpenClaw PR Closed: Passed All Reviews But Overwritten

Background

On May 11, I found Issue #80721 in OpenClaw project describing missing --dry-run support for config unset command. Based on that existing Issue, I submitted PR #80773 with implementation.

Timeline

Date	Event
May 11	Found existing Issue #80721, submitted PR #80773 based on it
May 11-13	ClawSweeper conducted 8 rounds of review, all P1/P2 issues fixed, passed all audits
May 13	Asked in PR if any improvements needed, expressed willingness to collaborate
May 13	Gio Della-Libera (Microsoft VP) submitted PR #81497
May 13	Gio closed #81497, admitting #80773 was "more complete"
May 14	Gio submitted PR #81895, re-implementing same feature, no communication
May 14	ClawSweeper Bot closed #80773, reason: #81895 is maintainer-labeled PR
May 15	Posted appeal comments in PR and Issue

Technical Comparison

Metric	PR #80773 (Independent)	PR #81895 (Maintainer)
Review Status	✅ Passed all 8 rounds	❌ Has 2 P2 defects
JSON Output	✅ Correct	❌ ClawSweeper flagged as defect
SecretRef Detection	✅ Complete paths	❌ ClawSweeper flagged as defect
Documentation	✅ Complete	❌ Missing
Co-author	-	❌ Not given

Key Fact: The two defects in #81895 were exactly what #80773 fixed in rounds 3-4.

Maintainer's Public Acknowledgment

When Gio closed his own PR #81497, he wrote:

"Closing this draft as a duplicate of #80773, which already covers config unset --dry-run more completely with docs, JSON output, behavior proof, and review history. We should let that PR be the canonical fix for #80721."

This comment is public record, visible to anyone.

Yet less than 24 hours later, the same person submitted a functionally overlapping PR #81895.

ClawSweeper's Position

In #81895 review, ClawSweeper acknowledged "overlapping implementation":

"Two narrow P2 dry-run defects remain and can be repaired mechanically if maintainers choose this PR as the canonical branch over the overlapping implementation."

When closing #80773, ClawSweeper updated:

"Close as superseded: the same config unset --dry-run work is now owned by the open maintainer-labeled PR at #81895."

Key: Any PR with maintainer label automatically becomes canonical implementation, regardless of submission order or technical quality.

Governance Opacity

According to CONTRIBUTING.md, official maintainer list has 28 people, but giodl73 (Gio Della-Libera) is not among them.

Yet PR #81895 has the maintainer label, and ClawSweeper used this as priority reason to close #80773.

Contradiction	Source
giodl73 not in maintainer list	CONTRIBUTING.md
PR #81895 has maintainer label	GitHub PR labels
ClawSweeper used this as priority reason	ClawSweeper comment

The maintainer label granting mechanism is opaque.

My Comments

Collaboration Inquiry (No Response)

Hi @giodl73-repo, I noticed PR #81895 also addresses #80721.
Is there anything I should improve or change in this PR (#80773)?
Happy to help if there's a way to combine efforts or if additional work is needed.

Expressing Disappointment

I'm disappointed that this PR was closed without any response to my comment
or explanation of the decision, especially after passing all ClawSweeper reviews.

Appeal to steipete

Hi @steipete, I'd like to bring your attention to this issue.

My PR #80773 passed all 8 rounds of ClawSweeper review and was explicitly
acknowledged by @giodl73-repo as "more complete" before he closed his own
draft #81497.

However, my PR was then closed by ClawSweeper Bot in favor of @giodl73-repo's
later PR #81895, which still has 2 P2 defects unresolved.

I would appreciate your review on this matter. Thank you.

Problem Summary

Issue	Description
First come first serve	Independent contributor submitted first, passed review first
Technical quality	Independent implementation more complete, Maintainer version has P2 issues
Communication missing	Maintainer didn't respond to any comments
Decision opaque	Bot closed PR citing maintainer-labeled, no explanation
Credit missing	No co-author given to independent contributor

Behavior Analysis

This is a classic "Not Invented Here" phenomenon, but more severe:

Maintainer already acknowledged external contribution quality in writing, still chose to re-implement
Entire process bypassed external contributor, no communication anywhere
Alternative implementation still has known defects that external version already fixed

If architectural direction was the sole consideration, maintainer should have commented on external PR requesting adjustments, not overruled without communication.

Project vs Individual Behavior

OpenClaw has 2000+ contributors, overall the project accepts external contributions.

This is giodl73's individual behavior, not project policy.

Level	Attitude
OpenClaw Project	Overall open source, accepts contributions
giodl73 Individual	This behavior doesn't respect contributors

Can't deny the whole project for one person's actions. But different maintainers have different styles.

My Takeaways

Despite PR being closed, technical contribution left public record:

Maintainer's written acknowledgment: PR #81497's close comment is objective fact visible to anyone
Complete technical record: 8 rounds of review, all fix commits are public proof of technical capability

Also: Learn maintainer's style before submitting PR LOL

Key Lessons

1. ClawSweeper is Iterative Review

Request re-review after each fix, updates comments in-place.

2. Real Behavior Proof Format is Strict

Must use - **Key**: format, must have real terminal output.

3. CI Failure May Not Be Your Problem

Check if it involves files you modified, if not, no action needed.

4. Documentation is Equally Important

P3 level findings also need attention.

5. Maintainers May Not Follow Open Source Spirit

Even if you:

Submitted first
Passed review
Expressed willingness to collaborate

Maintainer may still:

Open their own PR
Not communicate with you
Close your PR directly
Not give you co-author

This is reality. Learn and move on.