中文
~/home / Python / GitHub
GitHub · Python

harbor

harbor is a skill AI agent for Provide a unified standard to test an AI agent’s ability to complete tasks, maintain robustness, and ensure reproducibility during development. Pricing: not stated on the product page. As of 2026-10-25, KanonAgent records 1.0k upvotes. First indexed by KanonAgent on 2026-07-30.
Harbor is a framework that enables developers to evaluate and improve AI agents using standardized benchmarks for task completion, robustness, and reproducibility.
harbor — official preview image
1.0k upvotes
Tracked by Kanon since Jul 30, 2026
no signal yet
status
77/100
momentum · conf 0.92
63d
tracked since 2026-08-22
6
evidence records

No signal event on file yet — status is unknown, not "quiet".

📈 Timelinewhat changed, and when
🔎 Known / Unknown 12 of 23 fields unknown
Machine-callable— unknown —
Open source1 verified 1 quote(s)
Self-hostable1 inferred 1 quote(s)
Bring your own key1 inferred 1 quote(s)
Autonomy level2 inferred 2 quote(s)
Pricing model— unknown —
Integrationsdaytona,modal,langsmith,blaxel,novita inferred 1 quote(s)

“Unknown” means we have not verified it — it is not a “no”. Hard filters never treat unknown as false.

🤖 Agent teardown · skill
Job to be doneProvide a unified standard to test an AI agent’s ability to complete tasks, maintain robustness, and ensure reproducibility during development.
AutonomyL2 · tool-calling(evidence: "harbor run --dataset [email protected]…")
Who it is forAI researchers and agent developers
Prerequisitesopen source · self-hostable · bring your own API key
Integrationsdaytona · modal · langsmith · blaxel · novita
Traction · why it is rising861 PH votes on ProductHunt, indicating strong community interest in agent evaluation tool
Why it matters

It fills a critical gap in AI agent engineering by establishing a standardized evaluation layer for reliable, repeatable systems.

Evidence quotesverbatim, from the product’s own materials

“public GitHub repository (README fetched)”— structural
“harbor run --dataset [email protected] --agent claude-code”— readme
“export ANTHROPIC_API_KEY=<YOUR-KEY>”— readme
“harbor run --dataset [email protected]”— readme
“harbor run --agent claude-code --model”— readme
“providers like Daytona, Modal, LangSmith, Blaxel, and Novita Sandbox”— readme
Python
Signal source: GitHub
Visit official site →
📛 Official badgefor your site / README
harbor badge
Building harbor? Pick a style above — the embed code updates live. Deep color control via URL params: bg= / fg= / accent= (hex). It links back to this page.
Share on X
On mobile tap Share for WeChat / RED (Xiaohongshu) / X; on desktop use Copy text and paste into the app.

FAQ

What is harbor?

Harbor is a framework that enables developers to evaluate and improve AI agents using standardized benchmarks for task completion, robustness, and reproducibility.

What does harbor do?

Provide a unified standard to test an AI agent’s ability to complete tasks, maintain robustness, and ensure reproducibility during development.

Why does harbor matter?

It fills a critical gap in AI agent engineering by establishing a standardized evaluation layer for reliable, repeatable systems.

Is harbor open source?

Yes — harbor is open source.

Can I self-host harbor?

Yes — harbor can be self-hosted.

What does harbor integrate with?

daytona,modal,langsmith,blaxel,novita

How popular is harbor?

As tracked by KanonAgent: 1.0k upvotes (first indexed 2026-07-30).

What are the best harbor alternatives?

Similar AI agents tracked by KanonAgent: Stealth Venture, open-design, deer-flow, ruflo, career-ops, cherry-studio.

harbor alternatives — similar AI agents

Stealth Ventureopen-designdeer-flowruflocareer-opscherry-studio

Where this fits — browse the same shelf

AI Agent Skills & PluginsAI agents for Code development toolsAre there open-source AI agents for Code development tools?Can I self-host an AI agent for Code development tools?Self-hosted & open-source AI agents