Token导航 LogoToken导航

Agent Skill · 研究检索

pplx-sdk-dev(pplx SDK DEV 搜索)

pplx-sdk-dev 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

pplx-sdk-dev 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为操作浏览器;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 58,150。

总安装 420周安装 17GitHub Stars 公开资料未说明下载 132核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pplx-sdk-dev(pplx SDK DEV 搜索)
来源仓库:https://github.com/pv-udpv/pplx-sdk
仓库路径:skills/pplx-sdk-dev
安装命令:
npx skills add https://github.com/pv-udpv/pplx-sdk --skill pplx-sdk-dev
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/pv-udpv/pplx-sdk --skill pplx-sdk-dev

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词或任务场景快速定位候选结果时使用。适用于 pplx SDK DEV 搜索相关的信息查询与整理,可结合来源仓库进一步核验用法。

通过 npx skills add 命令从指定 GitHub 仓库安装,需确认权限范围和维护状态。安装前建议检查是否会触发联网、命令执行或文件读写,确保操作边界清晰。pplx-sdk-dev 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/pplx-sdk-dev

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/pv-udpv/pplx-sdk --skill pplx-sdk-dev;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要操作浏览器,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

pplx-sdk-dev 怎么安装?

在终端执行 npx skills add https://github.com/pv-udpv/pplx-sdk --skill pplx-sdk-dev。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

pplx-sdk-dev 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

pplx-sdk-dev 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

pplx-sdk-dev 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

pplx-sdk-dev 用的人多吗?

公开数据是累计安装 420 次,近一周安装 17 次,下载 132 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

pplx-sdk-dev 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
pplx-sdk-dev

pplx SDK DEV 搜索 · 本页

420—操作浏览器未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

pplx-sdk Development Meta-Skill

This meta-skill orchestrates all project-specific and community skills for pplx-sdk development. Activate this skill for any development task — it coordinates the right sub-skills automatically.

When to use

Use this skill for any development task on the pplx-sdk project: implementing features, fixing bugs, reviewing code, writing tests, or scaffolding new modules.

Subagent Architecture

Each skill delegates to a specialist subagent via context: fork. Subagents run in isolated context windows with restricted tool access.

┌───────────────────────────────────────────────────────┐
│                  orchestrator                         │
│  (meta-orchestrator, delegates subtasks)              │
├───────────────────────────────────────────────────────┤
│                                                       │
│  ┌──────────────┐  ┌──────────────┐                  │
│  │ code-reviewer │  │ test-runner   │                  │
│  │ (read-only)   │  │ (run & fix)   │                  │
│  │ view,grep,    │  │ bash,view,    │                  │
│  │ glob,bash     │  │ edit,grep     │                  │
│  └──────────────┘  └──────────────┘                  │
│                                                       │
│  ┌──────────────┐  ┌──────────────┐                  │
│  │ scaffolder    │  │ sse-expert    │                  │
│  │ (create new)  │  │ (streaming)   │                  │
│  │ view,edit,    │  │ view,edit,    │                  │
│  │ bash,grep     │  │ bash,grep     │                  │
│  └──────────────┘  └──────────────┘                  │
│                                                       │
│  ┌──────────────┐  ┌──────────────┐                  │
│  │ reverse-     │  │ architect     │                  │
│  │  engineer    │  │ (diagrams &   │                  │
│  │ (API disc.)  │  │  design)      │                  │
│  │ view,edit,   │  │ view,edit,    │                  │
│  │ bash,grep    │  │ bash,grep     │                  │
│  └──────────────┘  └──────────────┘                  │
│                                                       │
│  ┌──────────────┐  ┌──────────────┐                  │
│  │ spa-expert   │  │ codegraph     │                  │
│  │ (SPA RE,     │  │ (AST, deps,   │                  │
│  │  CDP, ext.)  │  │  knowledge)   │                  │
│  │ view,edit,   │  │ view,edit,    │                  │
│  │ bash,grep    │  │ bash,grep     │                  │
│  └──────────────┘  └──────────────┘                  │
│                                                       │
└───────────────────────────────────────────────────────┘

Subagent Definitions (.claude/agents/)

SubagentRoleToolsIsolation
orchestratorDecomposes tasks, coordinates othersview, bash, grep, globfork
code-reviewerRead-only architecture reviewview, grep, glob, bashfork
test-runnerRun tests, diagnose, fix failuresbash, view, edit, grep, globfork
scaffolderCreate new modules and filesview, edit, bash, grep, globfork
sse-expertSSE streaming implementationview, edit, bash, grep, globfork
reverse-engineerAPI discovery, traffic analysisview, edit, bash, grep, globfork
architectArchitecture diagrams, design validationview, edit, bash, grep, globfork
spa-expertSPA RE: React/Vite/Workbox/CDP, extensionsview, edit, bash, grep, globfork
codegraphAST parsing (Python + JS/TS), dep graphs, knowledge graphsview, edit, bash, grep, globfork

Skill Dependencies

This meta-skill composes the following skills. Apply them in the order shown based on the task type.

Project-Specific Skills (repo root)

SkillPathSubagentPurpose
code-reviewcode-review/SKILL.mdcode-reviewerArchitecture compliance and conventions
test-fixtest-fix/SKILL.mdtest-runnerDiagnose and fix failing tests
scaffold-modulescaffold-module/SKILL.mdscaffolderCreate new modules per layered architecture
sse-streamingsse-streaming/SKILL.mdsse-expertSSE protocol and streaming patterns
reverse-engineerreverse-engineer/SKILL.mdreverse-engineerAPI discovery from browser traffic
architecturearchitecture/SKILL.mdarchitectSystem diagrams and design visualization
spa-reverse-engineerspa-reverse-engineer/SKILL.mdspa-expertReact/Vite/Workbox/CDP webapp RE
code-analysiscode-analysis/SKILL.mdcodegraphAST parsing (Python + JS/TS), dependency graphs, knowledge graphs

Community Skills (installed via npx skills)

SkillSourcePurpose
python-testing-patternswshobson/agentspytest patterns, fixtures, mocking
python-type-safetywshobson/agentsType annotations, mypy strict mode
python-error-handlingwshobson/agentsException hierarchies, error patterns
python-design-patternswshobson/agentsProtocol pattern, DI, factory
api-design-principleswshobson/agentsREST API design, endpoint conventions
async-python-patternswshobson/agentsasync/await, concurrency patterns
skill-creatoranthropics/skillsGuide for creating new skills
mcp-builderanthropics/skillsBuild MCP servers for tool integration
webapp-testinganthropics/skillsTest web apps with Playwright
ast-grepast-grep/agent-skillStructural code search via AST patterns
knowledge-graph-builderdaffy0208/ai-dev-standardsKnowledge graph design and entity relationships
steeringnahisaho/codegraphmcpserverCodeGraph MCP orchestration and traceability

Workflow: New Feature

[plan] → [explore] → [research] → [implement] → [test] → [review] → [done]
  1. Plan — orchestrator decomposes the task, identifies target layer
  2. Explore — code-reviewer (read-only fork) analyzes existing code and dependencies
  3. Research — reverse-engineer or sse-expert (fork) studies API behavior, reads docs, gathers context needed before implementation
  4. Scaffold — scaffolder (fork) creates files from templates
  5. Implement — sse-expert or scaffolder (fork) writes the code
  6. Test — test-runner (fork) runs pytest, fixes failures
  7. Review — code-reviewer (fork) validates architecture compliance
  8. Verify — ruff check --fix. && ruff format. && mypy pplx_sdk/ && pytest -v

Workflow: Bug Fix

  1. Reproduce — test-runner (fork) runs failing test with -v
  2. Research — code-reviewer (fork) traces the code path, reads related modules and tests
  3. Diagnose — test-runner reads traceback, identifies root cause
  4. Fix — test-runner edits source/test, applies python-error-handling patterns
  5. Verify — test-runner runs full suite

Workflow: SSE/Streaming Work

  1. Research — sse-expert (fork) reviews SSE protocol spec, existing transport code, and API response patterns
  2. Implement — sse-expert writes streaming code, applies async-python-patterns
  3. Test — test-runner (fork) runs tests with mock SSE responses
  4. Review — code-reviewer (fork) validates transport layer compliance

Workflow: API Discovery (Reverse Engineering)

[capture] → [research] → [document] → [scaffold] → [test] → [review]
  1. Capture — reverse-engineer (fork) analyzes cURL/traffic from perplexity.ai DevTools
  2. Research — reverse-engineer studies request/response patterns, tests variations, maps edge cases and auth flows
  3. Document — reverse-engineer writes endpoint documentation with field types and examples
  4. Scaffold — scaffolder (fork) creates Pydantic models and service methods from schema
  5. Implement — sse-expert or scaffolder implements transport and domain code
  6. Test — test-runner (fork) validates with mock responses matching discovered schemas
  7. Review — code-reviewer (fork) ensures architecture compliance

Workflow: API Endpoint

  1. Research — code-reviewer (fork) studies existing endpoints and api-design-principles patterns
  2. Design — code-reviewer reviews proposed schema against REST conventions
  3. Implement — scaffolder (fork) creates FastAPI endpoint
  4. Test — test-runner (fork) validates with pytest-httpx mocks

Workflow: Architecture Design & Visualization

[analyze] → [research] → [diagram] → [validate] → [document]
  1. Analyze — architect (fork) reads existing code, maps imports and layer dependencies
  2. Research — architect studies the components involved and their relationships
  3. Diagram — architect produces Mermaid diagrams (layer map, sequence, class hierarchy, state machine)
  4. Validate — architect checks for circular deps, upward imports, protocol conformance
  5. Document — architect embeds diagrams in README, docs, or PR descriptions

Workflow: SPA Reverse Engineering

[detect] → [research] → [intercept] → [extract] → [code-graph] → [document] → [implement]
  1. Detect — spa-expert (fork) identifies the SPA stack (React version, bundler, state management, service workers)
  2. Research — spa-expert studies the SPA internals, React component tree, and network patterns
  3. Intercept — spa-expert captures network traffic via CDP, Chrome extension, or DevTools
  4. Extract — spa-expert extracts React state shapes, API schemas, and Workbox cache strategies
  5. Code Graph — codegraph (fork) analyzes SPA source code: component tree, import graph, hook chains, TypeScript types; cross-references with spa-expert runtime findings
  6. Document — reverse-engineer + spa-expert map runtime + static discoveries to SDK architecture
  7. Implement — scaffolder creates models and services; spa-expert builds tools/extensions

Workflow: Code Analysis & Knowledge Graph

[parse] → [graph] → [analyze] → [report] → [act]
  1. Parse — codegraph (fork) parses source AST: Python via ast module, JavaScript/TypeScript via grep-based import parsing
  2. Graph — codegraph builds dependency graph (module-to-module imports) and knowledge graph (Python: IMPORTS, INHERITS, IMPLEMENTS, CALLS, RAISES; SPA: RENDERS, USES_HOOK, CALLS_API, PROVIDES, CONSUMES)
  3. Analyze — codegraph detects patterns: circular deps, layer violations, dead code, complexity hotspots (Python); barrel file cycles, prop drilling, orphan routes (SPA)
  4. Report — codegraph produces structured insights report with Mermaid diagrams and actionable findings
  5. Act — Delegate findings: architect updates diagrams, code-reviewer reviews violations, scaffolder fixes gaps, spa-expert cross-references runtime analysis

Project Quick Reference

# Install dependencies
pip install -e ".[dev]"

# Install/update community skills
npx skills add wshobson/agents --skill python-testing-patterns --skill python-type-safety --skill python-error-handling --skill python-design-patterns --skill api-design-principles --skill async-python-patterns --agent github-copilot -y
npx skills add anthropics/skills --skill skill-creator --skill mcp-builder --skill webapp-testing --agent github-copilot -y
npx skills add ast-grep/agent-skill --skill ast-grep --agent github-copilot -y
npx skills add "daffy0208/ai-dev-standards@knowledge-graph-builder" --agent github-copilot -y
npx skills add nahisaho/codegraphmcpserver --skill steering --agent github-copilot -y

# Lint, format, type-check, test
ruff check --fix . && ruff format .
mypy pplx_sdk/ --ignore-missing-imports
pytest tests/ -v --cov=pplx_sdk

# Manage skills
npx skills list          # Show installed skills
npx skills check         # Check for updates
npx skills update        # Update all skills

Architecture Invariants

These rules must never be violated regardless of which sub-skill is active:

  1. Layer dependencies: core/ → shared/ → transport/ → domain/ → client.py (never reverse)
  2. Exception hierarchy: All errors extend PerplexitySDKError
  3. Type annotations: 100% coverage, from __future__ import annotations
  4. Google docstrings: On all public APIs
  5. Protocol pattern: Use typing.Protocol, never ABC

平台分布

Codex

37.34%
按下载量换算49

Claude

29.39%
按下载量换算39

Cursor

19.77%
按下载量换算26

Gemini CLI

8.92%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills