Terminal Bench Science AI Agent
by harbor-frameworkTerminal-Bench-Science: Evaluating AI agents on research workflows across scientific domains
Terminal Bench Science is an open-source ai agent by harbor-framework, written mainly in Python, with 798 GitHub stars and the Apache-2.0 license. Its code was last updated Nov 1, 2026.
What Terminal Bench Science can touch
- Not checked directly. Its own description suggests it does this.Run shell commandsLikely, from its own description
- This listing has not been scanned yet.Run dynamic codeNot scanned yet
- This listing has not been scanned yet.Read or write filesNot scanned yet
- This listing has not been scanned yet.Read environment and secretsNot scanned yet
- This listing has not been scanned yet.Control a web browserNot scanned yet
- This listing has not been scanned yet.Connect to databasesNot scanned yet
- This listing has not been scanned yet.Read or send messagesNot scanned yet
- This listing has not been scanned yet.Touch money or tradingNot scanned yet
- This listing has not been scanned yet.Connect to the internetNot scanned yet
"Not seen" means our checks did not find it, and the tool may still do it. See the scan policy.
Before you install Terminal Bench Science
Terminal Bench Science has not been scanned by AgenticMart yet. It can get a free automated scan. Until then, here is what it can likely access and what to check before you install it.
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 run shell commands.
Before you install
- Keep tool-call approval on, and run it in a container if you can.
- 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 harbor-framework on GitHub, 798 stars and 516 forks. Popularity is not proof of safety.
- Last code change: Nov 1, 2026.
- Open-source license: Apache-2.0.
Terminal Bench Science is an open-source ai agent by harbor-framework, written mainly in Python. Terminal-Bench-Science: Evaluating AI agents on research workflows across scientific domains
At a glance
- Stars
- 798
- Forks
- 516
- Language
- Python
- License
- Apache-2.0
- Created
- Jan 2026
- Last update
- Nov 1, 2026
- Type
- AI Agent
- Price
- Free
Topics
Questions about Terminal Bench Science
What is Terminal Bench Science?
Terminal-Bench-Science: Evaluating AI agents on research workflows across scientific domains
Who maintains Terminal Bench Science?
Terminal Bench Science is maintained by harbor-framework on GitHub (harbor-framework/terminal-bench-science). AgenticMart lists it from public repository data.
Is Terminal Bench Science safe to use?
Terminal Bench Science has not been scanned by AgenticMart yet. It can get a free automated scan. Until then, here is what it can likely access and what to check before you install it.
What license does Terminal Bench Science use?
It is published under the Apache-2.0 license.
Terminal Bench Science alternatives
View more →Similar ai agents in Coding Agents, ranked by GitHub stars. Compare licence, recent activity and scan status before you choose.
| Name | Stars | Licence | Last updated | Scan |
|---|---|---|---|---|
| Terminal Bench Science (this page) harbor-framework | 798 | Apache-2.0 | Nov 1, 2026 | Scanned, see report |
| AutoGPT Significant-Gravitas | 188k | None found | Nov 1, 2026 | Not scanned yet |
| Langchain langchain-ai | 149k | MIT | Nov 1, 2026 | Scanned, no flagged issues |
| Gemini CLI google-gemini | 109k | Apache-2.0 | Nov 1, 2026 | Scanned, no flagged issues |
| MetaGPT FoundationAgents | 71k | MIT | Jan 21, 2026 | Scanned, no flagged issues |
| Oh My Openagent code-yeongyu | 71k | None found | Nov 1, 2026 | Not scanned yet |