← 返回 Skills 市场
5005
总下载
4
收藏
8
当前安装
19
版本数
在 OpenClaw 中安装
/install tensorpm
功能描述
AI-powered project management - a Notion and Jira alternative with local-first architecture. Manage projects, track action items, and coordinate teams via MCP tools or A2A agent communication. Signed & notarized. https://tensorpm.com
安全使用建议
Install only if you trust the TensorPM desktop app and its download channels. If you use A2A, enable A2A_HTTP_AUTH_TOKEN before starting TensorPM, run it only on trusted machines, and avoid placing secrets in projects, imported files, or conversations unless you are comfortable with local agents and processes potentially accessing that data.
功能分析
Type: OpenClaw Skill
Name:
Developer:
Version:
Description: OpenClaw Agent Skill
The skill bundle describes an AI-powered project management application with local-first architecture. Installation instructions use standard package managers (Homebrew) or direct downloads from `tensorpm.com` and GitHub releases. The `SKILL.md` documents a local A2A agent endpoint running on `localhost:37850` without authentication, which is a common design for local services but means any local process can interact. It also details a feature to create projects from local files (`documentPath`). While these capabilities involve local network and file access, they are explicitly for the stated purpose of the application and do not show any clear evidence of intentional malicious behavior, data exfiltration, or prompt injection with harmful objectives from the skill bundle itself.
能力评估
Purpose & Capability
The MCP tools, A2A agent, project/action-item management, workspace switching, file-based project creation, and AI-provider key setup all fit the stated project-management purpose, but they touch sensitive business data and account credentials.
Instruction Scope
The documentation states that all localhost A2A requests are trusted by default, while the exposed API can list/read projects, read conversation history, create projects, create/update action items, and activate workspaces.
Install Mechanism
The skill artifact is documentation only and directs users to install an external desktop app through Homebrew, tensorpm.com, or GitHub release binaries; no executable code is bundled in the skill itself.
Credentials
A local service is coherent for a local-first app, but unauthenticated localhost access is broad for project-management data because other local processes may be able to interact with it.
Persistence & Privilege
Persistent project data, message history, workspace state, and write-only storage of AI provider keys are disclosed and purpose-aligned, but users should treat them as sensitive.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install tensorpm - 安装完成后，直接呼叫该 Skill 的名称或使用
/tensorpm触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.1.12
Reduced unnecessary overhead
v1.1.11
No user-visible changes in this release; documentation remains unchanged.
v1.1.10
- skill.md was renamed to SKILL.md (case change only).
- The SKILL metadata was updated with improved description, author information, and compatibility notes.
- No changes to functionality or tool interfaces; documentation has been clarified and standardized.
v1.1.9
No changes detected in this version.
- Version 1.1.9 was published with no file or documentation changes compared to the previous release.
v1.1.8
Version 1.1.8 of TensorPM
Added dependency functionality to for action items
v1.1.7
Version 1.1.7
- Added new modes to the create_project MCP tool: basic, fromPrompt, and fromFile.
- Updated MCP tool documentation to reflect detailed parameter information and new functionality.
- Enhanced documentation for action item fields.
- No code or functionality changes detected; updates are documentation-focused.
v1.1.6
New Skills for project creation
# Basic (instant)
curl -X POST http://localhost:37850/projects \
-H "Content-Type: application/json" \
-d '{"name": "New Project", "description": "Optional description"}'
# From prompt (AI-generated, async)
curl -X POST http://localhost:37850/projects \
-H "Content-Type: application/json" \
-d '{"name": "Mobile App", "mode": "fromPrompt", "prompt": "Build a habit tracker with streaks"}'
# From file (AI-generated from document, async)
curl -X POST http://localhost:37850/projects \
-H "Content-Type: application/json" \
-d '{"name": "From Brief", "mode": "fromFile", "documentPath": "/path/to/brief.pdf"}'
v1.1.5
- Updated description to position TensorPM as a Notion and Jira alternative.
- No functional changes; documentation and branding improvements only.
v1.1.4
TensorPM 1.1.4
- Updated Linux download instructions to use https://tensorpm.com/download/linux for easier access.
- No changes to features, tools, or API; documentation update only.
v1.1.3
- Updated Linux installation instructions: now uses direct AppImage download with `curl` instead of package-specific scripts.
- Minor formatting fix in the "Create Project" usage example (`description` field typo corrected).
v1.1.2
- Added detailed terminal-based installation instructions for Linux (Ubuntu/Debian and Fedora/RHEL) to the documentation.
- Improved clarity and completeness of the "Download" section to better support Linux users.
- No changes to skill logic or feature set; documentation update only.
v1.1.1
- Added information highlighting that macOS builds are code-signed and notarized by Apple.
- Noted that Windows builds are now signed via Azure Trusted Signing.
- Updated description to include "Signed & notarized."
- No functional or API changes; this release focuses on improved application security and trust.
v1.1.0
No changes detected in this release.
- Version 1.0.6 was published with no file or documentation updates.
- Functionality, documentation, and usage remain unchanged from the previous version.
v1.0.5
- Added Homebrew installation instructions for macOS.
- Added new workspace management features: list workspaces (`list_workspaces`) and switch active workspace (`set_active_workspace`) via MCP tools and REST API.
- Updated download section formatting for clarity and added additional download methods.
- Expanded available A2A REST endpoints to include workspace-related routes.
- Updated documentation for MCP tools to include new workspace commands and clarify results.
- Minor improvements and clarifications throughout setup and usage sections.
v1.0.4
- Added details about local-first approach with no account required and free usage.
- Noted support for both personal API keys and local models (Ollama, vLLM, LLM studio).
- Mentioned the new optional cloud sync with end-to-end encrypted collaboration.
- No changes detected in commands, endpoints, or tool usage.
v1.0.3
- Expanded A2A documentation with step-by-step agent discovery and communication examples.
- Updated JSON-RPC message format: message parts now use `"kind"` instead of `"type"`.
- Documented supported agent methods (`message/send`, `message/stream`, and task management methods).
- Added details on message part types, file attachments, and structured data in multi-part messages.
- Clarified MCP tool limitations and directed project structure changes to A2A agent requests.
- Improved overall documentation structure and examples for clarity.
v1.0.2
- Updated alternative download link from https://tensorpm.com/download to https://tensorpm.com
- No other changes detected in this version
v1.0.1
TensorPM 1.0.1
- Reformatted the download section for clarity, using concise list items with direct links.
- Updated links for release notes and alternative download to plain URLs.
- Trimmed introductory content for a more streamlined presentation—functional details remain unchanged.
v1.0.0
/Users/simon/Projects/TensorPM/SKILL.md
元数据
常见问题
AI-Driven Project Management: TensorPM 是什么？
AI-powered project management - a Notion and Jira alternative with local-first architecture. Manage projects, track action items, and coordinate teams via MCP tools or A2A agent communication. Signed & notarized. https://tensorpm.com. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 5005 次。
如何安装 AI-Driven Project Management: TensorPM？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install tensorpm」即可一键安装，无需额外配置。
AI-Driven Project Management: TensorPM 是免费的吗？
是的，AI-Driven Project Management: TensorPM 完全免费（开源免费），可自由下载、安装和使用。
AI-Driven Project Management: TensorPM 支持哪些平台？
AI-Driven Project Management: TensorPM 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 AI-Driven Project Management: TensorPM？
由 Neo552（@neo552）开发并维护，当前版本 v1.1.12。
推荐 Skills