# Harbor

## Docs

- [AgentContext](https://mintlify.wiki/harbor-framework/harbor/api/agent-context.md): Container for agent execution results and metadata
- [BaseAgent](https://mintlify.wiki/harbor-framework/harbor/api/base-agent.md): Abstract base class for implementing custom agents in Harbor
- [BaseEnvironment](https://mintlify.wiki/harbor-framework/harbor/api/base-environment.md): Abstract base class for containerized execution environments
- [Job](https://mintlify.wiki/harbor-framework/harbor/api/job.md): The main entrypoint for launching and managing Harbor evaluation trials
- [Job Configuration](https://mintlify.wiki/harbor-framework/harbor/api/job-config.md): Configuration models for Harbor jobs and orchestration
- [Job Results](https://mintlify.wiki/harbor-framework/harbor/api/job-result.md): Result models for Harbor job execution and statistics
- [Metrics](https://mintlify.wiki/harbor-framework/harbor/api/metrics.md): Metric configuration and base classes for computing evaluation metrics
- [Task Configuration](https://mintlify.wiki/harbor-framework/harbor/api/task-config.md): Configuration models for tasks, agents, environments, and verifiers
- [Trial Configuration](https://mintlify.wiki/harbor-framework/harbor/api/trial-config.md): Configuration for individual trial execution in Harbor
- [Trial Results](https://mintlify.wiki/harbor-framework/harbor/api/trial-result.md): Result models for individual trial execution in Harbor
- [harbor adapters](https://mintlify.wiki/harbor-framework/harbor/cli/adapters.md): Create and manage benchmark adapters
- [harbor cache](https://mintlify.wiki/harbor-framework/harbor/cli/cache.md): Manage Harbor cache and clean up resources
- [harbor datasets](https://mintlify.wiki/harbor-framework/harbor/cli/datasets.md): List and download datasets from registries
- [harbor jobs](https://mintlify.wiki/harbor-framework/harbor/cli/jobs.md): Manage and execute evaluation jobs
- [harbor run](https://mintlify.wiki/harbor-framework/harbor/cli/run.md): Start a job to evaluate agents on tasks
- [harbor sweeps](https://mintlify.wiki/harbor-framework/harbor/cli/sweeps.md): Run successive sweeps to focus on difficult tasks
- [harbor tasks](https://mintlify.wiki/harbor-framework/harbor/cli/tasks.md): Create and manage task definitions
- [harbor traces](https://mintlify.wiki/harbor-framework/harbor/cli/traces.md): Export agent trajectories to datasets
- [harbor trials](https://mintlify.wiki/harbor-framework/harbor/cli/trials.md): Run and manage individual trials
- [harbor view](https://mintlify.wiki/harbor-framework/harbor/cli/view.md): Interactive web UI for browsing agent trajectories
- [Agents](https://mintlify.wiki/harbor-framework/harbor/concepts/agents.md): Understanding the agent system and BaseAgent interface in Harbor
- [Environments](https://mintlify.wiki/harbor-framework/harbor/concepts/environments.md): Understanding execution environments in Harbor
- [Metrics](https://mintlify.wiki/harbor-framework/harbor/concepts/metrics.md): Understanding metrics collection and aggregation in Harbor
- [Tasks](https://mintlify.wiki/harbor-framework/harbor/concepts/tasks.md): Understanding task structure and components in Harbor
- [Trials and Jobs](https://mintlify.wiki/harbor-framework/harbor/concepts/trials-and-jobs.md): Understanding trial execution and job orchestration in Harbor
- [Verifiers](https://mintlify.wiki/harbor-framework/harbor/concepts/verifiers.md): Understanding the verification system in Harbor
- [Creating custom benchmarks](https://mintlify.wiki/harbor-framework/harbor/examples/custom-benchmark.md): Build your own evaluation benchmarks with Harbor's flexible task system
- [Implementing custom metrics](https://mintlify.wiki/harbor-framework/harbor/examples/custom-metrics.md): Create custom metrics to track and aggregate trial results in Harbor
- [Running parameter sweeps](https://mintlify.wiki/harbor-framework/harbor/examples/parameter-sweeps.md): Optimize agent configurations and hyperparameters using Harbor's sweep functionality
- [RL optimization with Harbor](https://mintlify.wiki/harbor-framework/harbor/examples/rl-optimization.md): Generate rollouts and trajectories for reinforcement learning optimization of AI agents
- [Running SWE-Bench evaluations](https://mintlify.wiki/harbor-framework/harbor/examples/swe-bench.md): Evaluate agents on real-world software engineering tasks from GitHub issues
- [Running Terminal-Bench evaluations](https://mintlify.wiki/harbor-framework/harbor/examples/terminal-bench.md): Learn how to evaluate agents on Terminal-Bench 2.0, the official benchmark for AI coding agents
- [Benchmark Adapters](https://mintlify.wiki/harbor-framework/harbor/guides/benchmark-adapters.md): Convert external benchmarks to Harbor task format
- [Cloud Execution](https://mintlify.wiki/harbor-framework/harbor/guides/cloud-execution.md): Run evaluations at scale using cloud providers like Daytona, Modal, and E2B
- [Creating Tasks](https://mintlify.wiki/harbor-framework/harbor/guides/creating-tasks.md): Build custom evaluation tasks for your agents
- [Custom Agents](https://mintlify.wiki/harbor-framework/harbor/guides/custom-agents.md): Implement your own agents to evaluate with Harbor
- [Parallel Execution](https://mintlify.wiki/harbor-framework/harbor/guides/parallel-execution.md): Optimize parallel execution strategies for large-scale evaluations
- [Running Evaluations](https://mintlify.wiki/harbor-framework/harbor/guides/running-evaluations.md): Learn how to run agent evaluations with Harbor
- [Installation](https://mintlify.wiki/harbor-framework/harbor/installation.md): Install Harbor and set up your evaluation environment
- [Supported Agents](https://mintlify.wiki/harbor-framework/harbor/integrations/agents.md): Complete list of AI agents supported by Harbor for evaluation
- [Supported Benchmarks](https://mintlify.wiki/harbor-framework/harbor/integrations/benchmarks.md): Complete list of benchmarks and datasets available in Harbor
- [Supported Environments](https://mintlify.wiki/harbor-framework/harbor/integrations/environments.md): Complete list of execution environments supported by Harbor
- [Introduction to Harbor](https://mintlify.wiki/harbor-framework/harbor/introduction.md): A powerful framework for evaluating and optimizing AI agents and language models in containerized environments
- [Quickstart](https://mintlify.wiki/harbor-framework/harbor/quickstart.md): Run your first Harbor evaluation in minutes