{
  "markdown": "# Arize Skills\n\nSkills that guide AI coding agents to help you add observability, run experiments, and optimize prompts for your LLM applications.\n\nThese skills encode the workflows we've refined building the [Arize](https://arize.com) platform and helping teams debug LLM apps in production. They handle the `ax` CLI flags, data shape quirks, and multi-step recipes so you don't have to.\n\nWorks with Cursor, Claude Code, Codex, GitHub Copilot, Windsurf, and [40+ other agents](https://github.com/vercel-labs/skills#supported-agents).\n\n## New to Arize? Start Here\n\n**Adding tracing to your app** — give your coding agent this prompt:\n\n> Install Arize skills from https://github.com/Arize-ai/arize-skills and use the arize-instrumentation skill to add tracing to this application.\n\nThe agent installs the skills, instruments your app, and verifies a trace reaches Arize AX — all in one session. You'll just need to provide your Arize credentials and approve the code changes.\n\n**Already have traces?** Give your agent this prompt to install the skills and start debugging:\n\n> Install Arize skills from https://github.com/Arize-ai/arize-skills, then use the arize-trace skill to export and analyze recent traces from my project. Summarize any errors or latency issues you find.\n\n## Installation\n\n### Option 1: npx (recommended)\n\nInstall all skills non-interactively — this is what the agent runs for the prompt above:\n\n```bash\nnpx skills add Arize-ai/arize-skills --skill \"*\" --yes\n```\n\nWant to hand-pick skills, agent, and scope yourself? Drop the flags for the interactive wizard:\n\n```bash\nnpx skills add Arize-ai/arize-skills\n```\n\nBoth auto-detect your agent (Cursor, Claude Code, Codex, etc.) and symlink skills into place.\n\n### Option 2: git clone\n\n**macOS / Linux:**\n```bash\ngit clone https://github.com/Arize-ai/arize-skills.git\ncd arize-skills\n./install.sh --project ~/my-project\n```\n\n**Windows (PowerShell):**\n```powershell\ngit clone https://github.com/Arize-ai/arize-skills.git\ncd arize-skills\n.\\install.ps1 -Project ~\\my-project\n```\n\nThe installer detects installed agents and optionally installs the `ax` CLI. Use `--global` / `-Global` instead to install to `~/.<agent>/skills/`.\n\n### Option 3: ax CLI\n\nIf you already have `ax` installed (v0.9.0+):\n\n```bash\nax skills install                                    # interactive\nax skills install --agent claude-code --yes         # non-interactive, Claude only\nax skills install --global --agent cursor --yes     # globally for Cursor\nax skills clear --yes                               # remove Arize skills\n```\n\nUse `--agent` (repeatable) to target specific agents: claude-code, cursor, codex, windsurf. Pass `--global` to install globally in `~/.<agents>/skills/` instead of your project, or `--yes` to skip confirmations.\n\n### Option 4: Claude Code plugin\n\nOpen up a Claude Code session and execute the following:\n\n```\n/plugin marketplace add Arize-ai/arize-skills\n/plugin install arize-skills@arize-skills\n```\n\n## Prerequisites\n\n### Arize CLI (`ax`)\n\nThe skills use the `ax` CLI to interact with the Arize API. Install it if you don't have it:\n\n```bash\n# Preferred (isolated environment)\nuv tool install arize-ax-cli\n# or\npipx install arize-ax-cli\n# Fallback\npip install arize-ax-cli\n```\n\n### Authentication\n\n**Option A — `ax` CLI profile** (recommended):\n\nSet up your API key once and it persists across all sessions and projects:\n```bash\n# Interactive wizard (creates 'default' profile if no profiles exist)\nax profiles create\n\n# Or pass the key directly (optional profile name as positional arg)\nax profiles create --api-key YOUR_API_KEY\nax profiles create staging --api-key YOUR_STAGING_KEY\n\n# Update an existing profile (patches only what you specify)\nax profiles update --api-key NEW_API_KEY\nax profiles update --region us-east-1b\n\n# Other profile management\nax profiles list\nax profiles show\nax profiles use staging\nax profiles delete staging\n```\n\nYou'll also need a space name or ID. Find yours by running `ax spaces list -o json` (use the `name` or base64 `id`), then persist it:\n```bash\n# macOS/Linux — add to ~/.zshrc or ~/.bashrc\nexport ARIZE_SPACE=\"my-workspace\"        # name, or base64 ID like U3BhY2U6...\n```\n\n**Option B — Environment variables**:\n```bash\nexport ARIZE_API_KEY=\"your-api-key\"       # from https://app.arize.com/admin > API Keys\nexport ARIZE_SPACE=\"my-workspace\"         # space name or base64 ID from ax spaces list\n# export ARIZE_DEFAULT_PROJECT=my-project # optional default project\n# export OPENAI_API_KEY=\"sk-...\"          # for AI integrations and evaluators\n# export ANTHROPIC_API_KEY=\"sk-ant-...\"   # for AI integrations and evaluators\n```\n\n### Verify\n\n```bash\nax --version && ax profiles show 2>&1\n```\n\n### On-prem / self-hosted Arize\n\nPoint the CLI at your deployment with a **single-endpoint** profile (hostname and HTTPS port, usually `443`). Replace the host with the value your operations team provides:\n\n```bash\nax profiles create my-onprem --api-key <key> --single-host arize.yourcompany.com --single-port 443\nax profiles use my-onprem\nax profiles validate\n```\n\nFor interactive setup, `ax profiles create` also offers **Advanced → Single endpoint**. More options (TOML, Flight/OTLP splits) are documented in the [arize-ax-cli README](https://github.com/Arize-ai/arize-ax-cli/blob/main/README.md#on-premise-deployments).\n\n## Available Skills\n\n| Skill | Description |\n|-------|-------------|\n| [arize-trace](skills/arize-trace/SKILL.md) | Export traces and spans by trace ID, span ID, or session ID. Debug LLM application issues. |\n| [arize-instrumentation](skills/arize-instrumentation/SKILL.md) | Add Arize AX tracing to an app. Two-phase flow: analyze codebase, then implement instrumentation (uses [Agent-Assisted Tracing](https://arize.com/docs/ax/alyx/tracing-assistant)). |\n| [arize-span-routing](skills/arize-span-routing/SKILL.md) | Send each agent's or tenant's Python spans to its correct Arize space and project using application metadata. |\n| [arize-dataset](skills/arize-dataset/SKILL.md) | Create, manage, and download datasets and examples. |\n| [arize-experiment](skills/arize-experiment/SKILL.md) | Run and analyze experiments against datasets. |\n| [arize-evaluator](skills/arize-evaluator/SKILL.md) | Create LLM-as-judge evaluators, run evaluation tasks, and set up continuous monitoring. |\n| [arize-ai-provider-integration](skills/arize-ai-provider-integration/SKILL.md) | Create and manage LLM provider credentials (OpenAI, Anthropic, Azure, Bedrock, Vertex, and more). |\n| [arize-annotation](skills/arize-annotation/SKILL.md) | Create and manage annotation configs (categorical, continuous, freeform); bulk-annotate project spans via the Python SDK. |\n| [arize-prompt-optimization](skills/arize-prompt-optimization/SKILL.md) | Optimize prompts using trace data, experiments, and meta-prompting. |\n| [arize-prompts](skills/arize-prompts/SKILL.md) | Manage Prompt Hub templates and versions with `ax prompts` (create, versions, labels). |\n| [arize-link](skills/arize-link/SKILL.md) | Generate deep links to traces, spans, and sessions in the Arize UI. |\n| [arize-compliance-audit](skills/arize-compliance-audit/SKILL.md) | Audit an AI agent for regulatory compliance (EU AI Act, NIST AI RMF, GDPR, HIPAA). Produces a tailored remediation checklist. |\n| [arize-admin](skills/arize-admin/SKILL.md) | Manage organizations, spaces, roles, role bindings, resource restrictions, and API keys. Enterprise admin workflows: team onboarding, SAML/SSO role mapping, service key management, and access control. |\n\n> [!WARNING]\n> **arize-compliance-audit is for guidance only and does not constitute legal advice or a complete compliance assessment.** It identifies common technical patterns based on publicly available regulatory frameworks and cannot account for your organisation's specific legal obligations, contractual commitments, or operational context. Always consult a qualified attorney or compliance specialist for binding assessments.\n\n## Installer Flags\n\n**Bash (`install.sh`):**\n\n| Flag | Description |\n|------|-------------|\n| `--project <dir>` | **Required.** Target project directory for skill symlinks |\n| `--global` | Install to `~/.<agent>/skills/` instead (alternative to `--project`) |\n| `--copy` | Copy files instead of symlinking |\n| `--force` | Overwrite existing skills |\n| `--skip-cli` | Don't install `ax` CLI even if missing |\n| `--agent <name>` | Manually specify agent (cursor, claude, codex, copilot) — repeatable |\n| `--skill <name>` | Only install/uninstall specific skills — repeatable (e.g. `--skill arize-trace --skill arize-dataset`) |\n| `--yes` | Skip confirmation prompts |\n| `--list` | List all available skills and exit |\n| `--uninstall` | Remove previously installed skill symlinks |\n\n**PowerShell (`install.ps1`):**\n\n| Flag | Description |\n|------|-------------|\n| `-Project <dir>` | **Required.** Target project directory for skill symlinks |\n| `-Global` | Install to `~/.<agent>/skills/` instead (alternative to `-Project`) |\n| `-Copy` | Copy files instead of symlinking |\n| `-Force` | Overwrite existing skills |\n| `-SkipCli` | Don't install `ax` CLI even if missing |\n| `-Agent <name>` | Manually specify agent (cursor, claude, codex, copilot) — repeatable |\n| `-Skill <name>` | Only install/uninstall specific skills — repeatable |\n| `-Yes` | Skip confirmation prompts |\n| `-Uninstall` | Remove previously installed skill symlinks |\n| `-List` | List all available skills and exit |\n\n## Updating\n\n- **npx path:** `npx skills update`\n- **git clone path:** `cd arize-skills && git pull` (symlinks update automatically)\n\n## Testing Skills\n\n`tests/run_skill.py` is an interactive test harness that runs a skill end-to-end using the Claude Agent SDK. It creates a temporary workspace, passes in your Arize credentials, and streams the agent's output.\n\nExport the required environment variables before running:\n\n```bash\nexport ARIZE_API_KEY=\"your-api-key\"       # from https://app.arize.com/admin > API Keys\nexport ARIZE_SPACE=\"my-workspace\"         # space name or base64 ID from ax spaces list\nexport TEST_PROJECT_NAME=\"my-project\"     # optional: Arize project for trace tests\nexport TEST_MODEL=\"claude-sonnet-4-6\"     # optional: Claude model override\n```\n\nThen run the harness:\n\n```bash\npython tests/run_skill.py --skill arize-trace --prompt \"Export trace abc123\"\n```\n\n> [!WARNING]\n> **Configure `.claude/settings.json` before running the test harness**\n>\n> The test harness uses Claude Code's `bypassPermissions` mode, which **skips all interactive\n> approval prompts**. This is safe because the agent runs in a sandboxed temporary workspace —\n> but **only** if your `settings.json` has a denylist blocking dangerous shell commands.\n>\n> **Without this, `bypassPermissions` gives the agent unrestricted shell access.**\n>\n> Add the following to `.claude/settings.json` in this repo (create it if it doesn't exist):\n>\n> ```json\n> {\n>   \"permissions\": {\n>     \"deny\": [\n>       \"Bash(rm -rf*)\",\n>       \"Bash(curl*)\",\n>       \"Bash(wget*)\",\n>       \"Bash(ssh*)\",\n>       \"Bash(scp*)\",\n>       \"Bash(git push*)\",\n>       \"Bash(sudo*)\",\n>       \"Bash(chmod*)\",\n>       \"Bash(chown*)\"\n>     ]\n>   }\n> }\n> ```\n\n## Links\n\n- [Arize Documentation](https://docs.arize.com)\n- [Arize REST API Reference](https://docs.arize.com/arize/api-reference/rest-api)\n- [ax CLI (arize-ax-cli)](https://github.com/Arize-ai/arize-ax-cli)\n- [OpenInference Semantic Conventions](https://github.com/Arize-ai/openinference)\n\n## License\n\nMIT\n",
  "bytes": 11473,
  "sha": "b9435866464cf12222f3b1499f8a7a7f6c61c515ad27348f52f6768f624c1796",
  "repo_slug": "arize-ai/arize-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/skl_arize_ai_arize_skills_arize_trace_ebbba259/readme"
}