{"kind":"Skill","metadata":{"namespace":"community","name":"arize-experiment","version":"0.1.0"},"spec":{"description":"Creates, runs, and analyzes Arize experiments for evaluating and comparing model performance. Covers experiment CRUD, exporting runs, comparing results, and evaluation workflows using the ax CLI. Use when the user mentions create experiment, run experiment, compare models, model performance, evaluate AI, experiment results, benchmark, A/B test models, or measure accuracy.","files":{"SKILL.md":"---\nname: arize-experiment\ndescription: Creates, runs, and analyzes Arize experiments for evaluating and comparing model performance. Covers experiment CRUD, exporting runs, comparing results, and evaluation workflows using the ax CLI. Use when the user mentions create experiment, run experiment, compare models, model performance, evaluate AI, experiment results, benchmark, A/B test models, or measure accuracy.\nmetadata:\n  author: arize\n  version: \"1.0\"\ncompatibility: Requires the ax CLI and a configured Arize profile.\n---\n\n# Arize Experiment Skill\n\n\u003e **`SPACE`** — All `--space` flags and the `ARIZE_SPACE` env var accept a space **name** (e.g., `my-workspace`) or a base64 space **ID** (e.g., `U3BhY2U6...`). Find yours with `ax spaces list`.\n\n## Concepts\n\n- **Experiment** = a named evaluation run against a specific dataset version, containing one run per example\n- **Experiment Run** = the result of processing one dataset example -- includes the model output, optional evaluations, and optional metadata\n- **Dataset** = a versioned collection of examples; every experiment is tied to a dataset and a specific dataset version\n- **Evaluation** = a named metric attached to a run (e.g., `correctness`, `relevance`), with optional label, score, and explanation\n\nThe typical flow: export a dataset → process each example → collect outputs and evaluations → create an experiment with the runs.\n\n## Prerequisites\n\nProceed directly with the task — run the `ax` command you need. Do NOT check versions, env vars, or profiles upfront.\n\nIf an `ax` command fails, troubleshoot based on the error:\n- `command not found` or version error → see references/ax-setup.md\n- `401 Unauthorized` / missing API key → run `ax profiles show` to inspect the current profile. If the profile is missing or the API key is wrong, follow references/ax-profiles.md to create/update it. If the user doesn't have their key, direct them to https://app.arize.com/admin \u003e API Keys\n- Space unknown → run `ax spaces list` to pick by name, or ask the user\n- Project unclear → ask the user, or run `ax projects list -o json --limit 100` and present as selectable options\n- **Security:** Never read `.env` files or search the filesystem for credentials. Use `ax profiles` for Arize credentials and `ax ai-integrations` for LLM provider keys. If credentials are not available through these channels, ask the user.\n- **CRITICAL — Never fabricate outputs:** When running an experiment, you MUST call the real model API specified by the user for every dataset example. Never fabricate, simulate, or hardcode model outputs, latencies, or evaluation scores. If you cannot call the API (missing SDK, missing credentials, network error), stop and tell the user what is needed before proceeding.\n\n## List Experiments: `ax experiments list`\n\nBrowse experiments, optionally filtered by dataset. Output goes to stdout.\n\n```bash\nax experiments list\nax experiments list --dataset DATASET_NAME --space SPACE --limit 20   # DATASET_NAME: name or ID (name preferred)\nax experiments list --cursor CURSOR_TOKEN\nax experiments list -o json\n```\n\n### Flags\n\n| Flag | Type | Default | Description |\n|------|------|---------|-------------|\n| `--dataset` | string | none | Filter by dataset |\n| `--limit, -l` | int | 15 | Max results (1-100) |\n| `--cursor` | string | none | Pagination cursor from previous response |\n| `-o, --output` | string | table | Output format: table, json, csv, parquet, or file path |\n| `-p, --profile` | string | default | Configuration profile |\n\n## Get Experiment: `ax experiments get`\n\nQuick metadata lookup -- returns experiment name, linked dataset/version, and timestamps.\n\n```bash\nax experiments get NAME_OR_ID\nax experiments get NAME_OR_ID -o json\nax experiments get NAME_OR_ID --dataset DATASET_NAME --space SPACE   # required when using experiment name instead of ID\n```\n\n### Flags\n\n| Flag | Type | Default | Description |\n|------|------|---------|-------------|\n| `NAME_OR_ID` | string | required | Experiment name or ID (positional) |\n| `--dataset` | string | none | Dataset name or ID (required if using experiment name instead of ID) |\n| `--space` | string | none | Space name or ID (required if using dataset name instead of ID) |\n| `-o, --output` | string | table | Output format |\n| `-p, --profile` | string | default | Configuration profile |\n\n### Response fields\n\n| Field | Type | Description |\n|-------|------|-------------|\n| `id` | string | Experiment ID |\n| `name` | string | Experiment name |\n| `dataset_id` | string | Linked dataset ID |\n| `dataset_version_id` | string | Specific dataset version used |\n| `experiment_traces_project_id` | string | Project where experiment traces are stored |\n| `created_at` | datetime | When the experiment was created |\n| `updated_at` | datetime | Last modification time |\n\n## Export Experiment: `ax experiments export`\n\nDownload all runs to a file. By default uses the REST API; pass `--all` to use Arrow Flight for bulk transfer.\n\n```bash\n# EXPERIMENT_NAME, DATASET_NAME: name or ID (name preferred)\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE\n# -\u003e experiment_abc123_20260305_141500/runs.json\n\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --all\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --output-dir ./results\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout | jq '.[0]'\n```\n\n### Flags\n\n| Flag | Type | Default | Description |\n|------|------|---------|-------------|\n| `NAME_OR_ID` | string | required | Experiment name or ID (positional) |\n| `--dataset` | string | none | Dataset name or ID (required if using experiment name instead of ID) |\n| `--space` | string | none | Space name or ID (required if using dataset name instead of ID) |\n| `--all` | bool | false | Use Arrow Flight for bulk export (see below) |\n| `--output-dir` | string | `.` | Output directory |\n| `--stdout` | bool | false | Print JSON to stdout instead of file |\n| `-p, --profile` | string | default | Configuration profile |\n\n### REST vs Flight (`--all`)\n\n- **REST** (default): Lower friction -- no Arrow/Flight dependency, standard HTTPS ports, works through any corporate proxy or firewall. Limited to 500 runs per page.\n- **Flight** (`--all`): Required for experiments with more than 500 runs. Uses gRPC+TLS on a separate host/port (`flight.arize.com:443`) which some corporate networks may block.\n\n**Agent auto-escalation rule:** If a REST export returns exactly 500 runs, the result is likely truncated. Re-run with `--all` to get the full dataset.\n\nOutput is a JSON array of run objects:\n\n```json\n[\n  {\n    \"id\": \"run_001\",\n    \"example_id\": \"ex_001\",\n    \"output\": \"The answer is 4.\",\n    \"evaluations\": {\n      \"correctness\": { \"label\": \"correct\", \"score\": 1.0 },\n      \"relevance\": { \"score\": 0.95, \"explanation\": \"Directly answers the question\" }\n    },\n    \"metadata\": { \"model\": \"gpt-4o\", \"latency_ms\": 1234 }\n  }\n]\n```\n\n## Create Experiment: `ax experiments create`\n\nCreate a new experiment with runs from a data file.\n\n```bash\nax experiments create --name \"gpt-4o-baseline\" --dataset DATASET_NAME --space SPACE --file runs.json\nax experiments create --name \"claude-test\" --dataset DATASET_NAME --space SPACE --file runs.csv\n```\n\n### Flags\n\n| Flag | Type | Required | Description |\n|------|------|----------|-------------|\n| `--name, -n` | string | yes | Experiment name |\n| `--dataset` | string | yes | Dataset to run the experiment against |\n| `--space, -s` | string | no | Space name or ID (required if using dataset name instead of ID) |\n| `--file, -f` | path | yes | Data file with runs: CSV, JSON, JSONL, or Parquet |\n| `-o, --output` | string | no | Output format |\n| `-p, --profile` | string | no | Configuration profile |\n\n### Passing data via stdin\n\nUse `--file -` to pipe data directly — no temp file needed:\n\n```bash\necho '[{\"example_id\": \"ex_001\", \"output\": \"Paris\"}]' | ax experiments create --name \"my-experiment\" --dataset DATASET_NAME --space SPACE --file -\n\n# Or with a heredoc\nax experiments create --name \"my-experiment\" --dataset DATASET_NAME --space SPACE --file - \u003c\u003c 'EOF'\n[{\"example_id\": \"ex_001\", \"output\": \"Paris\"}]\nEOF\n```\n\n### Required columns in the runs file\n\n| Column | Type | Required | Description |\n|--------|------|----------|-------------|\n| `example_id` | string | yes | ID of the dataset example this run corresponds to |\n| `output` | string | yes | The model/system output for this example |\n\nAdditional columns are passed through as `additionalProperties` on the run.\n\n## Delete Experiment: `ax experiments delete`\n\n```bash\nax experiments delete NAME_OR_ID\nax experiments delete NAME_OR_ID --dataset DATASET_NAME --space SPACE   # required when using experiment name instead of ID\nax experiments delete NAME_OR_ID --force   # skip confirmation prompt\n```\n\n### Flags\n\n| Flag | Type | Default | Description |\n|------|------|---------|-------------|\n| `NAME_OR_ID` | string | required | Experiment name or ID (positional) |\n| `--dataset` | string | none | Dataset name or ID (required if using experiment name instead of ID) |\n| `--space` | string | none | Space name or ID (required if using dataset name instead of ID) |\n| `--force, -f` | bool | false | Skip confirmation prompt |\n| `-p, --profile` | string | default | Configuration profile |\n\n## Experiment Run Schema\n\nEach run corresponds to one dataset example:\n\n```json\n{\n  \"example_id\": \"required -- links to dataset example\",\n  \"output\": \"required -- the model/system output for this example\",\n  \"evaluations\": {\n    \"metric_name\": {\n      \"label\": \"optional string label (e.g., 'correct', 'incorrect')\",\n      \"score\": \"optional numeric score (e.g., 0.95)\",\n      \"explanation\": \"optional freeform text\"\n    }\n  },\n  \"metadata\": {\n    \"model\": \"gpt-4o\",\n    \"temperature\": 0.7,\n    \"latency_ms\": 1234\n  }\n}\n```\n\n### Evaluation fields\n\n| Field | Type | Required | Description |\n|-------|------|----------|-------------|\n| `label` | string | no | Categorical classification (e.g., `correct`, `incorrect`, `partial`) |\n| `score` | number | no | Numeric quality score (e.g., 0.0 - 1.0) |\n| `explanation` | string | no | Freeform reasoning for the evaluation |\n\nAt least one of `label`, `score`, or `explanation` should be present per evaluation.\n\n## Workflows\n\n### Run an experiment against a dataset\n\n1. Find or create a dataset:\n   ```bash\n   ax datasets list --space SPACE\n   ax datasets export DATASET_NAME --space SPACE --stdout | jq 'length'\n   ```\n2. Export the dataset examples:\n   ```bash\n   ax datasets export DATASET_NAME --space SPACE\n   ```\n3. Call the real model API for each example and collect outputs. Use `ax datasets export --stdout` to pipe examples directly into an inference script:\n\n   ```bash\n   ax datasets export DATASET_NAME --space SPACE --stdout | python3 infer.py \u003e runs.json\n   ```\n\n   Write `infer.py` to read examples from stdin, call the target model, and write runs JSON to stdout. The script below is a template — first inspect the exported dataset JSON to find the correct input field name, then uncomment the provider block the user wants:\n\n   ```python\n   import json, sys, time\n\n   examples = json.load(sys.stdin)\n   runs = []\n\n   for ex in examples:\n       # Inspect the exported JSON to find the right field (e.g. \"input\", \"question\", \"prompt\")\n       user_input = ex.get(\"input\") or ex.get(\"question\") or ex.get(\"prompt\") or str(ex)\n\n       start = time.time()\n\n       # === CALL THE REAL MODEL API HERE — never fabricate or simulate ===\n       # Uncomment and adapt the provider block the user requested:\n       #\n       # OpenAI (pip install openai  — uses OPENAI_API_KEY env var):\n       #   from openai import OpenAI\n       #   resp = OpenAI().chat.completions.create(\n       #       model=\"gpt-4o\",\n       #       messages=[{\"role\": \"user\", \"content\": user_input}]\n       #   )\n       #   output_text = resp.choices[0].message.content\n       #\n       # Anthropic (pip install anthropic  — uses ANTHROPIC_API_KEY env var):\n       #   import anthropic\n       #   resp = anthropic.Anthropic().messages.create(\n       #       model=\"claude-sonnet-4-6\", max_tokens=1024,\n       #       messages=[{\"role\": \"user\", \"content\": user_input}]\n       #   )\n       #   output_text = resp.content[0].text\n       #\n       # Google Gemini (pip install google-genai  — uses GOOGLE_API_KEY env var):\n       #   from google import genai\n       #   resp = genai.Client().models.generate_content(\n       #       model=\"gemini-2.5-pro\", contents=user_input\n       #   )\n       #   output_text = resp.text\n       #\n       # Custom / OpenAI-compatible proxy (pip install openai — uses CUSTOM_BASE_URL + CUSTOM_API_KEY env vars):\n       # Use this for Azure OpenAI, NVIDIA NIM, local Ollama, or any OpenAI-compatible endpoint,\n       # including a test integration proxy. Matches the `custom` provider in `ax ai-integrations create`.\n       #   import os\n       #   from openai import OpenAI\n       #   resp = OpenAI(\n       #       base_url=os.environ[\"CUSTOM_BASE_URL\"],          # e.g. https://my-proxy.example.com/v1\n       #       api_key=os.environ.get(\"CUSTOM_API_KEY\", \"none\"),\n       #   ).chat.completions.create(\n       #       model=os.environ.get(\"CUSTOM_MODEL\", \"default\"),\n       #       messages=[{\"role\": \"user\", \"content\": user_input}]\n       #   )\n       #   output_text = resp.choices[0].message.content\n\n       latency_ms = round((time.time() - start) * 1000)\n       runs.append({\n           \"example_id\": ex[\"id\"],\n           \"output\": output_text,\n           \"metadata\": {\"model\": \"MODEL_NAME\", \"latency_ms\": latency_ms}\n       })\n       print(f\"  {ex['id']}: {latency_ms}ms\", file=sys.stderr)\n\n   json.dump(runs, sys.stdout, indent=2)\n   ```\n\n   **Before running:** install the provider SDK (`pip install openai` / `anthropic` / `google-genai`) and ensure the API key is set as an environment variable in your shell. If you cannot access the API, stop and tell the user what is needed.\n\n4. Verify the runs file:\n   ```bash\n   python3 -c \"import json; runs=json.load(open('runs.json')); print(f'{len(runs)} runs'); print(json.dumps(runs[0], indent=2))\"\n   ```\n   Each run must have `example_id` and `output`. Optional fields: `evaluations`, `metadata`.\n5. Create the experiment:\n   ```bash\n   ax experiments create --name \"gpt-4o-baseline\" --dataset DATASET_NAME --space SPACE --file runs.json\n   ```\n6. Verify: `ax experiments get \"gpt-4o-baseline\" --dataset DATASET_NAME --space SPACE`\n\n### Compare two experiments\n\n1. Export both experiments:\n   ```bash\n   ax experiments export \"experiment-a\" --dataset DATASET_NAME --space SPACE --stdout \u003e a.json\n   ax experiments export \"experiment-b\" --dataset DATASET_NAME --space SPACE --stdout \u003e b.json\n   ```\n2. Compare evaluation scores by `example_id`:\n   ```bash\n   # Average correctness score for experiment A\n   jq '[.[] | .evaluations.correctness.score] | add / length' a.json\n\n   # Same for experiment B\n   jq '[.[] | .evaluations.correctness.score] | add / length' b.json\n   ```\n3. Find examples where results differ:\n   ```bash\n   jq -s '.[0] as $a | .[1][] | . as $run |\n     {\n       example_id: $run.example_id,\n       b_score: $run.evaluations.correctness.score,\n       a_score: ($a[] | select(.example_id == $run.example_id) | .evaluations.correctness.score)\n     }' a.json b.json\n   ```\n4. Score distribution per evaluator (pass/fail/partial counts):\n   ```bash\n   # Count by label for experiment A\n   jq '[.[] | .evaluations.correctness.label] | group_by(.) | map({label: .[0], count: length})' a.json\n   ```\n5. Find regressions (examples that passed in A but fail in B):\n   ```bash\n   jq -s '\n     [.[0][] | select(.evaluations.correctness.label == \"correct\")] as $passed_a |\n     [.[1][] | select(.evaluations.correctness.label != \"correct\") |\n       select(.example_id as $id | $passed_a | any(.example_id == $id))\n     ]\n   ' a.json b.json\n   ```\n\n**Statistical significance note:** Score comparisons are most reliable with ≥ 30 examples per evaluator. With fewer examples, treat the delta as directional only — a 5% difference on n=10 may be noise. Report sample size alongside scores: `jq 'length' a.json`.\n\n### Download experiment results for analysis\n\n1. `ax experiments list --dataset DATASET_NAME --space SPACE` -- find experiments\n2. `ax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE` -- download to file\n3. Parse: `jq '.[] | {example_id, score: .evaluations.correctness.score}' experiment_*/runs.json`\n\n### Pipe export to other tools\n\n```bash\n# Count runs\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout | jq 'length'\n\n# Extract all outputs\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout | jq '.[].output'\n\n# Get runs with low scores\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout | jq '[.[] | select(.evaluations.correctness.score \u003c 0.5)]'\n\n# Convert to CSV\nax experiments export EXPERIMENT_NAME --dataset DATASET_NAME --space SPACE --stdout | jq -r '.[] | [.example_id, .output, .evaluations.correctness.score] | @csv'\n```\n\n## Related Skills\n\n- **arize-dataset**: Create or export the dataset this experiment runs against → use `arize-dataset` first\n- **arize-prompt-optimization**: Use experiment results to improve prompts → next step is `arize-prompt-optimization`\n- **arize-trace**: Inspect individual span traces for failing experiment runs → use `arize-trace`\n- **arize-link**: Generate clickable UI links to traces from experiment runs → use `arize-link`\n\n## Troubleshooting\n\n| Problem | Solution |\n|---------|----------|\n| `ax: command not found` | See references/ax-setup.md |\n| `401 Unauthorized` | API key is wrong, expired, or doesn't have access to this space. Fix the profile using references/ax-profiles.md. |\n| `No profile found` | No profile is configured. See references/ax-profiles.md to create one. |\n| `Experiment not found` | Verify experiment name with `ax experiments list --space SPACE` |\n| `Invalid runs file` | Each run must have `example_id` and `output` fields |\n| `example_id mismatch` | Ensure `example_id` values match IDs from the dataset (export dataset to verify) |\n| `No runs found` | Export returned empty -- verify experiment has runs via `ax experiments get` |\n| `Dataset not found` | The linked dataset may have been deleted; check with `ax datasets list` |\n\n## Save Credentials for Future Use\n\nSee references/ax-profiles.md § Save Credentials for Future Use.\n","references/ax-profiles.md":"# ax Profile Setup\n\nConsult this when authentication fails (401, missing profile, missing API key). Do NOT run these checks proactively.\n\nUse this when there is no profile, or a profile has incorrect settings (wrong API key, wrong region, etc.).\n\n## 1. Inspect the current state\n\n```bash\nax profiles show\n```\n\nLook at the output to understand what's configured:\n- `API Key: (not set)` or missing → key needs to be created/updated\n- No profile output or \"No profiles found\" → no profile exists yet\n- Connected but getting `401 Unauthorized` → key is wrong or expired\n- Connected but wrong endpoint/region → region needs to be updated\n\n## 2. Fix a misconfigured profile\n\nIf a profile exists but one or more settings are wrong, patch only what's broken.\n\n**Never pass a raw API key value as a flag.** Always reference it via the `ARIZE_API_KEY` environment variable. If the variable is not already set in the shell, instruct the user to set it first, then run the command:\n\n```bash\n# If ARIZE_API_KEY is already exported in the shell:\nax profiles update --api-key $ARIZE_API_KEY\n\n# Fix the region (no secret involved — safe to run directly)\nax profiles update --region us-east-1b\n\n# Fix both at once\nax profiles update --api-key $ARIZE_API_KEY --region us-east-1b\n```\n\n`update` only changes the fields you specify — all other settings are preserved. If no profile name is given, the active profile is updated.\n\n## 3. Create a new profile\n\nIf no profile exists, or if the existing profile needs to point to a completely different setup (different org, different region):\n\n**Always reference the key via `$ARIZE_API_KEY`, never inline a raw value.**\n\n```bash\n# Requires ARIZE_API_KEY to be exported in the shell first\nax profiles create --api-key $ARIZE_API_KEY\n\n# Create with a region\nax profiles create --api-key $ARIZE_API_KEY --region us-east-1b\n\n# Create a named profile\nax profiles create work --api-key $ARIZE_API_KEY --region us-east-1b\n```\n\nTo use a named profile with any `ax` command, add `-p NAME`:\n```bash\nax spans export PROJECT -p work\n```\n\n## 4. Getting the API key\n\n**Never ask the user to paste their API key into the chat. Never log, echo, or display an API key value.**\n\nIf `ARIZE_API_KEY` is not already set, instruct the user to export it in their shell:\n\n```bash\nexport ARIZE_API_KEY=\"...\"   # user pastes their key here in their own terminal\n```\n\nThey can find their key at https://app.arize.com/admin \u003e API Keys. Recommend they create a **scoped service key** (not a personal user key) — service keys are not tied to an individual account and are safer for programmatic use. Keys are space-scoped — make sure they copy the key for the correct space.\n\nOnce the user confirms the variable is set, proceed with `ax profiles create --api-key $ARIZE_API_KEY` or `ax profiles update --api-key $ARIZE_API_KEY` as described above.\n\n## 5. Verify\n\nAfter any create or update:\n\n```bash\nax profiles show\n```\n\nConfirm the API key and region are correct, then retry the original command.\n\n## Space\n\nThere is no profile flag for space. Save it as an environment variable — accepts a space **name** (e.g., `my-workspace`) or a base64 space **ID** (e.g., `U3BhY2U6...`). Find yours with `ax spaces list -o json`.\n\n**macOS/Linux** — add to `~/.zshrc` or `~/.bashrc`:\n```bash\nexport ARIZE_SPACE=\"my-workspace\"    # name or base64 ID\n```\nThen `source ~/.zshrc` (or restart terminal).\n\n**Windows (PowerShell):**\n```powershell\n[System.Environment]::SetEnvironmentVariable('ARIZE_SPACE', 'my-workspace', 'User')\n```\nRestart terminal for it to take effect.\n\n## Save Credentials for Future Use\n\nAt the **end of the session**, if the user manually provided any credentials during this conversation **and** those values were NOT already loaded from a saved profile or environment variable, offer to save them.\n\n**Skip this entirely if:**\n- The API key was already loaded from an existing profile or `ARIZE_API_KEY` env var\n- The space was already set via `ARIZE_SPACE` env var\n- The user only used base64 project IDs (no space was needed)\n\n**How to offer:** Use **AskQuestion**: *\"Would you like to save your Arize credentials so you don't have to enter them next time?\"* with options `\"Yes, save them\"` / `\"No thanks\"`.\n\n**If the user says yes:**\n\n1. **API key** — Run `ax profiles show` to check the current state. Then run `ax profiles create --api-key $ARIZE_API_KEY` or `ax profiles update --api-key $ARIZE_API_KEY` (the key must already be exported as an env var — never pass a raw key value).\n\n2. **Space** — See the Space section above to persist it as an environment variable.\n","references/ax-setup.md":"# ax CLI — Troubleshooting\n\nConsult this only when an `ax` command fails. Do NOT run these checks proactively.\n\n## Check version first\n\nIf `ax` is installed (not `command not found`), always run `ax --version` before investigating further. The version must be `0.14.0` or higher — many errors are caused by an outdated install. If the version is too old, see **Version too old** below.\n\n## `ax: command not found`\n\n**macOS/Linux:**\n1. Check common locations: `~/.local/bin/ax`, `~/Library/Python/*/bin/ax`\n2. Install: `uv tool install arize-ax-cli` (preferred), `pipx install arize-ax-cli`, or `pip install arize-ax-cli`\n3. Add to PATH if needed: `export PATH=\"$HOME/.local/bin:$PATH\"`\n\n**Windows (PowerShell):**\n1. Check: `Get-Command ax` or `where.exe ax`\n2. Common locations: `%APPDATA%\\Python\\Scripts\\ax.exe`, `%LOCALAPPDATA%\\Programs\\Python\\Python*\\Scripts\\ax.exe`\n3. Install: `pip install arize-ax-cli`\n4. Add to PATH: `$env:PATH = \"$env:APPDATA\\Python\\Scripts;$env:PATH\"`\n\n## Version too old (below 0.14.0)\n\nUpgrade: `uv tool install --force --reinstall arize-ax-cli`, `pipx upgrade arize-ax-cli`, or `pip install --upgrade arize-ax-cli`\n\n## SSL/certificate error\n\n- macOS: `export SSL_CERT_FILE=/etc/ssl/cert.pem`\n- Linux: `export SSL_CERT_FILE=/etc/ssl/certs/ca-certificates.crt`\n- Fallback: `export SSL_CERT_FILE=$(python -c \"import certifi; print(certifi.where())\")`\n\n## Subcommand not recognized\n\nUpgrade ax (see above) or use the closest available alternative.\n\n## Still failing\n\nStop and ask the user for help.\n"},"import":{"commit_sha":"541b7819d8c3545c6df122491af4fa1eae415779","imported_at":"2026-05-18T20:05:35Z","license_text":"MIT License\n\nCopyright GitHub, Inc.\n\nPermission is hereby granted, free of charge, to any person obtaining a copy\nof this software and associated documentation files (the \"Software\"), to deal\nin the Software without restriction, including without limitation the rights\nto use, copy, modify, merge, publish, distribute, sublicense, and/or sell\ncopies of the Software, and to permit persons to whom the Software is\nfurnished to do so, subject to the following conditions:\n\nThe above copyright notice and this permission notice shall be included in all\ncopies or substantial portions of the Software.\n\nTHE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\nIMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\nFITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE\nAUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\nLIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\nOUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE\nSOFTWARE.","owner":"github","repo":"github/awesome-copilot","source_url":"https://github.com/github/awesome-copilot/tree/541b7819d8c3545c6df122491af4fa1eae415779/plugins/arize-ax/skills/arize-experiment"}},"content_hash":[5,143,192,111,23,132,160,146,162,205,175,150,159,156,206,221,132,171,170,153,87,39,218,161,223,24,248,247,229,16,108,194],"trust_level":"unsigned","yanked":false}