OpenAI-BackendTools
A collection of tools to test, operate and maintain OpenAI backends (OpenAI and AzureOpenAI)
Install / Use
npx skills add karstenheld3/OpenAI-BackendToolsInstalls into whichever agent you are using.
Devin Config
Devin AI agent config
Quality Score
Category
Development & EngineeringSupported Platforms
Our assessment of OpenAI-BackendTools
OpenAI-BackendTools scores 66/100 on our quality scale, 3948th of 4,642 Development & Engineering skills we index.
Its Devin Config is 6.1 KB long, well organised into 8 sections with 4 code examples: a thorough specification that gives an agent plenty to work with.
It has no GitHub stars yet, so there is no community track record; judge it on its content.
Maintenance, license and trust
- We could not determine when the repository was last updated.
- Our last check on 2026-09-28 found the source still online.
- No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
- Its trust signals score 68/100, with 3 cautions from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-09-24. Automated pattern scan on 2026-09-24. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
OpenAI-BackendTools compared with similar skills
All 4 of these similar skills score higher than OpenAI-BackendTools; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| OpenAI-BackendTools (this skill)by karstenheld3 | 66 | 0 | — | Devin Config |
| Agent-Reachby Panniantong | 100 | 90.5k | 19d ago | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 74.4k | today | CLAUDE.md |
| CowAgentby zhayujie | 100 | 47.2k | today | CLAUDE.md |
| ai-job-searchby MadsLorentzen | 100 | 44.9k | today | CLAUDE.md |
Frequently asked questions
- How do I install OpenAI-BackendTools?
- Run
npx skills add karstenheld3/OpenAI-BackendTools. The install tabs above show the steps for each supported agent. - Which AI agents does OpenAI-BackendTools work with?
- It is written for Devin, as a Devin Config file. Other agents that read the same format can often use it too.
- Is OpenAI-BackendTools safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It declares no license and scores 68/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is OpenAI-BackendTools still maintained?
- We could not determine when the repository was last updated.
Skill content
View source on GitHubPlaywright MCP Configuration Reference
Package: @playwright/mcp | Repo: github.com/microsoft/playwright-mcp | Latest: v0.0.70
Configuration Precedence
CLI flags > environment variables > JSON config file
Environment Variables
Every CLI flag has a PLAYWRIGHT_MCP_ env var equivalent. Uppercase, hyphens become underscores:
--no-sandbox=PLAYWRIGHT_MCP_NO_SANDBOX--user-data-dir=PLAYWRIGHT_MCP_USER_DATA_DIR--caps=PLAYWRIGHT_MCP_CAPS
CLI Flags Reference
Browser Selection and Launch:
--browser <browser>- chromium (default), chrome, firefox, webkit, msedge--executable-path <path>- Custom browser binary--headless- Headless mode (headed by default)--device <device>- Device emulation, e.g., "iPhone 15"--viewport-size <size>- Viewport, e.g., "1280x720"--user-agent <ua>- Custom user agent string
Profile and Session:
--user-data-dir <path>- Custom profile directory (overrides default workspace-scoped location)--isolated- In-memory profile, nothing persists--storage-state <path>- Load cookies/localStorage from JSON file--save-session- Save session to output directory
Initialization:
--init-page <path...>- TypeScript file evaluated on Playwright page object before automation--init-script <path...>- JavaScript injected before page scripts (multiple files allowed)
Capabilities:
--caps <caps>- Comma-separated opt-in categories: vision, pdf, devtools, network, storage, testing, config
Connection:
--extension- Connect via "Playwright MCP Bridge" Chrome extension--cdp-endpoint <endpoint>- Chrome DevTools Protocol endpoint--cdp-header <headers...>- CDP headers--cdp-timeout <timeout>- CDP connect timeout (default 30000ms)--endpoint <endpoint>- Bound browser endpoint
Server (HTTP/Server-Sent Events (SSE) mode):
--port <port>- Enable HTTP/SSE transport--host <host>- Bind address (default localhost, use 0.0.0.0 for all interfaces)--allowed-hosts <hosts...>- Allowed hosts ('*' disables check)--shared-browser-context- Share browser context between HTTP clients
Network:
--allowed-origins <origins>- Semicolon-separated trusted origins (NOT a security boundary, NOT affecting redirects)--blocked-origins <origins>- Semicolon-separated blocked origins (evaluated before allowlist)--block-service-workers- Block service workers--proxy-server <proxy>- HTTP proxy, e.g., "http://myproxy:3128"--proxy-bypass <bypass>- Domains to bypass proxy--ignore-https-errors- Ignore HTTPS certificate errors
Security:
--no-sandbox- Disable Chromium sandbox (required for Docker/WSL2, cosmetic warning elsewhere)--sandbox- Enable extra sandbox for normally non-sandboxed processes--secrets <path>- Dotenv file; matching text in responses is replaced (convenience, not secure)--allow-unrestricted-file-access- Allow file access outside workspace roots--grant-permissions <perms...>- Browser permissions: geolocation, clipboard-read, clipboard-write
Output:
--output-dir <path>- Directory for output files (screenshots, PDFs, videos)--output-mode <mode>- "file" or "stdout" (default)--image-responses <mode>- "allow", "omit", or "auto" (default)--snapshot-mode <mode>- "full" (default) or "none"--console-level <level>- Console level: error, warning, info, debug
Code Generation:
--codegen <lang>- "typescript" (default) or "none"
Testing:
--test-id-attribute <attr>- Test ID attribute (default "data-testid")
Timeouts:
--timeout-action <ms>- Action timeout (default 5000ms)--timeout-navigation <ms>- Navigation timeout (default 60000ms)
JSON Config File
Activated with --config path/to/config.json:
{
"browser": {
"browserName": "chromium",
"isolated": false,
"userDataDir": "/path/to/profile",
"launchOptions": {},
"contextOptions": {},
"cdpEndpoint": "ws://...",
"cdpHeaders": {},
"cdpTimeout": 30000,
"remoteEndpoint": "ws://...",
"initPage": ["init-page.ts"],
"initScript": ["init-script.js"]
},
"extension": false,
"server": {
"port": 8931,
"host": "localhost",
"allowedHosts": ["*"]
},
"capabilities": ["core", "vision", "pdf", "devtools", "network", "storage", "testing", "config"],
"saveSession": false,
"sharedBrowserContext": false,
"secrets": { "API_KEY": "sk-..." },
"outputDir": "./output",
"console": { "level": "info" },
"network": {
"allowedOrigins": ["https://example.com:*"],
"blockedOrigins": ["https://ads.example.com:*"]
},
"testIdAttribute": "data-testid",
"timeouts": { "action": 5000, "navigation": 60000, "expect": 5000 },
"imageResponses": "auto",
"snapshot": { "mode": "full" },
"allowUnrestrictedFileAccess": false,
"codegen": "typescript"
}
Default Profile Locations
Workspace-scoped persistent profiles (default, not isolated):
- Windows:
%LOCALAPPDATA%\ms-playwright\mcp-{channel}-{workspace-hash} - macOS:
~/Library/Caches/ms-playwright/mcp-{channel}-{workspace-hash} - Linux:
~/.cache/ms-playwright/mcp-{channel}-{workspace-hash}
Workspace hash derived from MCP client's workspace root - different projects get separate profiles. Override with --user-data-dir <path>. Delete profile directory to clear all state.
Deployment Modes
Stdio (default) - MCP client spawns server:
{"mcpServers": {"playwright": {"command": "npx", "args": ["@playwright/mcp@latest"]}}}
HTTP/SSE - Standalone server (for headed browser on headless systems):
npx @playwright/mcp@latest --port 8931
Client: {"mcpServers": {"playwright": {"url": "http://localhost:8931/mcp"}}}
Docker - Headless Chromium, --no-sandbox required:
{"mcpServers": {"playwright": {"command": "docker", "args": ["run", "-i", "--rm", "--init", "--pull=always", "mcr.microsoft.com/playwright/mcp"]}}}
Requirements
- Node.js 18+ with npx in PATH
- Chrome/Chromium for headed mode
- For extension mode: Chrome or Edge with "Playwright MCP Bridge" extension from Chrome Web Store
Related Skills
Agent-Reach
90.5kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
74.4kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
CowAgent
47.2kOpen-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
ai-job-search
44.9kThe job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep interviews. Fork it and own it.
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
