HarnessBench AI Agent
by reacher-zBenchmark for comparing agent harnesses on everyday online tasks — fixes the base model, varies the harness. Sister project of ClawBench, same scoring pipeline.
HarnessBench is an open-source ai agent by reacher-z, written mainly in Python, with 64 GitHub stars and the Apache-2.0 license. Its code was last updated Jul 17, 2026.
Scan summary: scanned Nov 5, 2026 at commit d7ff125. No capabilities flagged by our checks. Scan policy
What HarnessBench can touch
- Our checks did not find this. The tool may still do it.Run shell commandsNot seen in this scan
- Our checks did not find this. The tool may still do it.Run dynamic codeNot seen in this scan
- Our checks did not find this. The tool may still do it.Read or write filesNot seen in this scan
- Our checks did not find this. The tool may still do it.Read environment and secretsNot seen in this scan
- Not checked directly. Its own description suggests it does this.Control a web browserLikely, from its own description
- Our checks did not find this. The tool may still do it.Connect to databasesNot seen in this scan
- Our checks did not find this. The tool may still do it.Read or send messagesNot seen in this scan
- Our checks did not find this. The tool may still do it.Touch money or tradingNot seen in this scan
- Our checks did not find this. The tool may still do it.Connect to the internetNot run in this scan
"Not seen" means our checks did not find it, and the tool may still do it. Covers commit d7ff125, Nov 5, 2026. See the scan policy.
Before you install HarnessBench
AgenticMart scanned HarnessBench on Nov 5, 2026 with automated checks for hidden code, known-malicious packages, leaked secrets and vulnerable dependencies. The result covers one commit under our scan policy, so review what access you give it and check any advisories shown here.
What it can access
- An AI agent decides its own next steps and calls whatever tools you give it, so its risk depends on the access you grant.
- Based on its description and topics, it likely needs to control a web browser (visit sites, fill in and submit forms).
Before you install
- Run it in a separate browser profile without your logged-in accounts.
- Pin the version you reviewed instead of always installing the latest release.
- Read the tool list or scripts before the first run, and keep approval prompts on for actions that write, send or delete.
- Maintained by reacher-z on GitHub, 64 stars and 6 forks. Popularity is not proof of safety.
- Last code change: Jul 17, 2026.
- Open-source license: Apache-2.0.
HarnessBench is an open-source ai agent by reacher-z, written mainly in Python. Benchmark for comparing agent harnesses on everyday online tasks — fixes the base model, varies the harness. Sister project of ClawBench, same scoring pipeline.
At a glance
- Stars
- 64
- Forks
- 6
- Language
- Python
- License
- Apache-2.0
- Created
- Apr 2026
- Last update
- Jul 17, 2026
- Type
- AI Agent
- Price
- Free
Topics
Questions about HarnessBench
What is HarnessBench?
Benchmark for comparing agent harnesses on everyday online tasks — fixes the base model, varies the harness. Sister project of ClawBench, same scoring pipeline.
Who maintains HarnessBench?
HarnessBench is maintained by reacher-z on GitHub (reacher-z/HarnessBench). AgenticMart lists it from public repository data.
Is HarnessBench safe to use?
AgenticMart scanned HarnessBench on Nov 5, 2026 with automated checks for hidden code, known-malicious packages, leaked secrets and vulnerable dependencies. The result covers one commit under our scan policy, so review what access you give it and check any advisories shown here.
What license does HarnessBench use?
It is published under the Apache-2.0 license.
HarnessBench alternatives
View more →Similar ai agents in Productivity & Workflow, ranked by GitHub stars. Compare licence, recent activity and scan status before you choose.
| Name | Stars | Licence | Last updated | Scan |
|---|---|---|---|---|
| HarnessBench (this page) reacher-z | 64 | Apache-2.0 | Jul 17, 2026 | Scanned, no flagged issues |
| Paperclip paperclipai | 110k | MIT | Nov 1, 2026 | Scanned, no flagged issues |
| Ppt Master hugohe3 | 72k | MIT | Oct 30, 2026 | Scanned, no flagged issues |
| Llama Index run-llama | 53k | MIT | Oct 26, 2026 | Scanned, see report |
| Nanobot HKUDS | 50k | MIT | Oct 31, 2026 | Scanned, no flagged issues |
| Sim simstudioai | 30k | Apache-2.0 | Nov 1, 2026 | Scanned, see report |