AI robot hacked GitHub: 253 spam comments on an issue, the maintainer marked it with --author

AI bots are taking over the GitHub open source repository: a $900 bounty issue was filled with 253 useless comments, and the 27 PRs received were all untested. Maintainers use Git --author tags and self-built reputation bots to fight back against AI spam submissions.

Core conclusion

AI bots spamming commits on GitHub are becoming a nightmare for open source maintainers. After the Archestra team posted the $900 bounty, an issue was flooded with 253 comments by an AI bot, and it received 27 untested PRs within a week. The maintainer spent half a day cleaning up "AI garbage" every week. They ended up using Git’s --author tag and custom reputation bots to combat this AI spam war.

Key Points

Event time: Continuous fermentation from April to May 2026 -Affected objects: All open source project maintainers, especially projects with bounties
Core changes: AI code generation reduces the contribution threshold to zero, but the quality threshold is forced to rise

Background and trigger events

In April 2026, the Archestra (enterprise MCP platform) team released a $900 bounty task on GitHub, hoping that community contributors would add "MCP Apps" support to the platform. Soon, real contributors appeared - someone proposed a solution, someone submitted an attempt. But AI robots are coming too.

"AI accounts started pouring in, not just for this one issue - but for the entire warehouse." Archestra CTO Ildar Iskhakov wrote.

Within a week, this issue received 253 comments, most of which were AI-generated "implementation plan" text garbage. Under the same repository, a simple "add x.ai support" issue received 27 PR, most of which were not even tested.

This isn't just Archestra's case. GitHub officials previously shared data to celebrate AI's "huge contribution" to product metrics, but completely ignored the reality that the quality of contributions has seriously declined. Maintainers are forced to spend half a day every week cleaning up garbage, otherwise the repository becomes completely unattractive to real contributors.

Dimensions	Change	What it means to us	Recommended actions
Maintenance cost	The team cleans up rubbish PR half a day every week	Time is continuously consumed	Introducing an automated review process
Community health	Real contributors are being submerged	Open source ecosystem is deteriorating	Deploying a reputation scoring system
Tool response	Git `--author` tag + custom robot	Passive defense is a feasible solution	Use n8n to build an automatic PR filter
AI profit	Spam PR is identified and intercepted	Affects automated workflow	Optimize AI output quality threshold

Adaptation suggestions

1. Add quality gate to AI Agent workflow

If you are running an AI automated code generation pipeline, make sure the output has quality gates. Refer to the on-site guide: How to add quality gates to your AI automation workflow: A practical guide from output to trustworthy results for the method of "trustworthy output" - self-check before receiving an order and verification after output.

2. Use the --author flag to intercept AI contributions

Archestra’s approach is simple: use the --author tag in Git commit records to identify and filter AI-generated commits. In practical applications, you can combine GitHub Actions to automatically tag and tag AI-generated PRs.

3. Establish a reputation system

Archestra built a tiny robot called "London-Cat" that calculates contributor reputation based on merged PRs and other signals. This idea can be used for any automated workflow - instead of unconditionally trusting the source, scoring is based on historical performance.

with your automated workflows

The automatic code generation brought by AI Agent is a double-edged sword. The more pipes you have, the more quality guardrails you need. A complete practical guide is available on the site:

**Want to learn to build an automated quality gate? ** How to add quality gates to your AI automation workflow: A practical guide from output to trustworthy results teaches you how to add verification on the output side
Real Example: Indie Developer: n8n + OpenClaw Automation Workflow Earning $5,000/mo shows how to mass-produce content while maintaining quality

Tool entry

Tool entries that appear naturally in the text: Git, GitHub, n8n, OpenClaw, Claude, ChatGPT

Reference sources

Next action

If you are running an AI-automated content production pipeline, it is recommended to add quality gates today. The lower the threshold, the more guardrails are needed.