Developer toolchain（开发者工具链）¶
The Palantir platform comes with a set of developer tools that enable you to build on top of the Ontology and unlock value with new workflows, applications, and functionality. With Palantir's developer toolchain, you can access the full power of the Ontology directly from your development environment.
In addition to the data connection and application building tools described elsewhere, the Palantir developer toolchain provides:
- Core APIs and SDKs that open up the Ontology to your applications.
- Development environments and developer-centric functionality to accelerate your workflows.
- Palantir MCP to enable AI IDEs and AI agents to use Palantir context and take action throughout Foundry.
- Compute modules to containerize your code and scale your deployments.
Core APIs and SDKs¶
Central to Palantir's developer offerings are the APIs and SDKs that open up the platform and allow you to build applications that integrate directly with the Palantir platform and your Ontology data.
Ontology SDK (OSDK)¶
The Ontology SDK (OSDK) enables developers to generate SDKs from their ontologies in Python, Java, and TypeScript. OSDK lets you access object types, apply actions to update data in the Ontology, call functions, and run AIP Logic functions for AIP-enabled enrollments. The in-platform Developer Console includes Ontology-specific documentation for the entities chosen for your application. Applications use the OAuth flow as a public or confidential client to access the data.
APIs¶
Palantir's APIs support users building on the Palantir platform by enabling programmatic management of platform access and the data backing your Ontology. API references are also available in-platform via the Developer Console.
Available Palantir APIs include:
- Datasets — Read, write, and manage datasets, including branches, transactions, files, and views.
- Filesystem — Manage spaces, projects, folders, and resource roles.
- Administration — Manage users, groups, organizations, hosts, and markings.
- Orchestration — Trigger builds, inspect jobs, and configure schedules.
- Connectivity — Manage connections, virtual tables, and file or table imports from external sources.
- SQL Queries — Execute SQL against datasets.
- Media Sets [Beta] — Upload, retrieve, and manage media sets and their items.
See the full API documentation for the complete list of endpoints, including models, audit logs, data health checks, streams, checkpoints, and Notepad.
Platform SDKs¶
Palantir's platform SDKs are publicly available for Python ↗ and TypeScript ↗, and are expanded as new APIs become available. These SDKs are usable alongside Ontology SDKs with the same Palantir platform clients.
Developer environments and functionality¶
In addition to development tools focused on data connection such as Pipeline Builder and Code Repositories, Code Workspaces allows you to use familiar IDEs like VS Code while building on top of the Palantir platform. VS Code workspaces are integrated with Developer Console, allowing you to rapidly build React applications. You can create a VS Code workspace from the in-platform Developer Console.
Palantir's developer-focused functionality aims to accelerate the process of building on the platform. Features like Global Branching help create a safe, streamlined environment for development of end-to-end workflows.
Palantir MCP¶
Palantir MCP is an implementation of the Model Context Protocol ↗ that allows AI IDEs and AI agents to autonomously build end-to-end applications in the Palantir platform - from data integration through ontology configuration and application development. In addition, you can use Palantir MCP to allow external AI systems to query documentation, metadata, and data, as well as perform high-level tasks on the platform. Developers can use Palantir MCP to automate auxiliary tasks while they stay focused on the system they are building.
Compute Modules¶
The Compute Modules feature enables you to deploy interactive containers on the Palantir platform, allowing you to bring in your existing code base (regardless of language) and run it inside the platform. With Compute Modules, for example, you could bring a third-party machine learning model into the platform and integrate it into your workflow.
For more information about Compute Modules, you can watch this official Build with AIP: Compute Modules ↗ video and reference the documentation.
Custom endpoints¶
Custom endpoints enable developers to configure and deploy user-defined API endpoints with their own URL patterns, request and response shapes, and specifications, while leveraging Foundry's back-end capabilities. These endpoints are backed by the ontology through actions and functions.
中文翻译¶
开发者工具链¶
Palantir 平台提供了一套开发者工具，使您能够在本体（Ontology）之上进行构建，并通过新的工作流、应用程序和功能释放价值。借助 Palantir 的开发者工具链，您可以直接从开发环境中访问本体的全部能力。
除了其他地方描述的数据连接和应用程序构建工具外，Palantir 开发者工具链还提供：
- 核心 API 和 SDK，将本体开放给您的应用程序。
- 开发环境和面向开发者的功能，以加速您的工作流。
- Palantir MCP，使 AI IDE 和 AI 代理能够使用 Palantir 上下文并在整个 Foundry 中执行操作。
- 计算模块，用于容器化您的代码并扩展您的部署。
核心 API 和 SDK¶
Palantir 开发者产品的核心是 API 和 SDK，它们开放了平台，允许您构建与 Palantir 平台和本体数据直接集成的应用程序。
本体 SDK (OSDK)¶
本体 SDK (OSDK) 使开发者能够从其本体生成 Python、Java 和 TypeScript 的 SDK。OSDK 允许您访问对象类型（object types）、应用操作（actions）以更新本体中的数据、调用函数（functions），并为已启用 AIP 的注册环境运行 AIP Logic 函数。平台内的开发者控制台（Developer Console）包含针对您应用程序所选实体的本体特定文档。应用程序使用 OAuth 流程作为公共或机密客户端来访问数据。
API¶
Palantir 的 API 支持用户在 Palantir 平台上进行构建，通过实现对平台访问权限和本体底层数据的编程管理。API 参考也可通过平台内的开发者控制台获取。
可用的 Palantir API 包括：
- 数据集 — 读取、写入和管理数据集，包括分支、事务、文件和视图。
- 文件系统 — 管理空间、项目、文件夹和资源角色。
- 管理 — 管理用户、组、组织、主机和标记。
- 编排 — 触发构建、检查作业和配置调度。
- 连接 — 管理连接、虚拟表以及来自外部源的文件或表导入。
- SQL 查询 — 对数据集执行 SQL。
- 媒体集 [Beta] — 上传、检索和管理媒体集及其项目。
请参阅完整的 API 文档，获取包括模型、审计日志、数据健康检查、流、检查点和 Notepad 在内的完整端点列表。
平台 SDK¶
Palantir 的平台 SDK 可公开获取，适用于 Python ↗ 和 TypeScript ↗，并随着新 API 的推出而扩展。这些 SDK 可与本体 SDK 一起使用，共享相同的 Palantir 平台客户端。
开发环境和功能¶
除了专注于数据连接的开发工具（如 Pipeline Builder 和 代码仓库）之外，代码工作区 允许您在 Palantir 平台上进行构建时使用熟悉的 IDE（如 VS Code）。VS Code 工作区 与开发者控制台集成，使您能够快速构建 React 应用程序。您可以从平台内的开发者控制台创建 VS Code 工作区。
Palantir 面向开发者的功能旨在加速在平台上的构建过程。诸如全局分支 等功能有助于为端到端工作流的开发创建一个安全、精简的环境。
Palantir MCP¶
Palantir MCP 是模型上下文协议 ↗ 的一种实现，允许 AI IDE 和 AI 代理在 Palantir 平台上自主构建端到端应用程序——从数据集成到本体配置和应用程序开发。此外，您可以使用 Palantir MCP 允许外部 AI 系统查询文档、元数据和数据，以及在平台上执行高级任务。开发者可以使用 Palantir MCP 来自动化辅助任务，同时专注于他们正在构建的系统。
计算模块¶
计算模块（Compute Modules）功能使您能够在 Palantir 平台上部署交互式容器，允许您引入现有代码库（无论使用何种语言）并在平台内运行。例如，使用计算模块，您可以将第三方机器学习模型引入平台并将其集成到您的工作流中。
有关计算模块的更多信息，您可以观看此官方 Build with AIP: Compute Modules ↗ 视频并参考文档。
自定义端点¶
自定义端点 使开发者能够配置和部署用户定义的 API 端点，这些端点拥有自己的 URL 模式、请求和响应形状以及规范，同时利用 Foundry 的后端能力。这些端点通过操作和函数由本体支持。