{
  "markdown": "# Oodle Agent Skills\n\nPublic repository of agent skills for the [Oodle](https://oodle.ai) observability platform. Each skill is a structured `SKILL.md` document that teaches AI coding assistants (Claude Code, Cursor, Codex, GitHub Copilot, Factory.ai, Gemini CLI, Windsurf, and the Oodle CLI itself) how to drive the [`oodle` CLI](https://github.com/oodle-ai/oodle-cli) correctly — with prescriptive rules, executable examples, and clear failure-handling guidance.\n\nThese skills are designed so that an AI assistant can:\n\n- Discover and operate Oodle resources (monitors, dashboards, notifiers, traces, metrics, synthetic monitors, drop rules, log metrics, integrations) and run PromQL/log queries with a single, predictable workflow.\n- Avoid common mistakes such as alert fatigue, accidental dashboard overwrites, untested notifiers, and runaway metric ingestion costs.\n- Produce output that other tools and scripts can consume (`-o json`) and run safely in CI (`--force`, `--retries N`).\n\n## Available Skills\n\n| Skill | Description |\n|-------|-------------|\n| [oodle-cli](skills/oodle-cli/SKILL.md) | Core CLI usage — auth, output formats, time flags, file input, common patterns |\n| [oodle-monitors](skills/oodle-monitors/SKILL.md) | Monitor CRUD, alerting thresholds, query scoping, runbook links |\n| [oodle-dashboards](skills/oodle-dashboards/SKILL.md) | Dashboard and folder management with safe-update / safe-delete patterns |\n| [oodle-alerting](skills/oodle-alerting/SKILL.md) | Notifiers, notification policies, muting rules |\n| [oodle-metrics](skills/oodle-metrics/SKILL.md) | Metric queries, label discovery, PromQL building |\n| [oodle-traces](skills/oodle-traces/SKILL.md) | Trace search by service, operation, duration, error |\n| [oodle-log-metrics](skills/oodle-log-metrics/SKILL.md) | Log-based metric rules — filters and groupBy |\n| [oodle-synthetic](skills/oodle-synthetic/SKILL.md) | HTTP/TCP synthetic monitors with assertions |\n| [oodle-metrics-query](skills/oodle-metrics-query/SKILL.md) | PromQL instant and range queries |\n| [oodle-logs](skills/oodle-logs/SKILL.md) | Log search and index pattern discovery |\n| [oodle-drop-rules](skills/oodle-drop-rules/SKILL.md) | Metric drop / sample rules for ingestion cost control |\n| [oodle-onboarding](skills/oodle-onboarding/SKILL.md) | Integration onboarding — list integrations, fetch setup specs, execute step-by-step installation |\n| [oodle-o11y-context](skills/oodle-o11y-context/SKILL.md) | Change-context manifest — the PR block that carries a change's observability intent across the SDLC |\n| [oodle-o11y-review](skills/oodle-o11y-review/SKILL.md) | Left-side telemetry review — business context, cardinality, duplication, RED/USE/fan-out, alert-driven scrutiny |\n| [oodle-triage](skills/oodle-triage/SKILL.md) | Right-side triage — single-alert context gathering and windowed auto-triage with fingerprint dedup |\n\n## Install\n\n### Native plugin marketplaces (recommended — auto-updating)\n\nSeveral agents can install this repo as a plugin straight from GitHub and **auto-update** it — no file copying, no manual re-pulls. In every case the marketplace is named `oodle-ai` and the plugin is `oodle-agent-skills`, so the install target is `oodle-agent-skills@oodle-ai`.\n\n**Claude Code** — reads `.claude-plugin/marketplace.json`:\n\n```\n/plugin marketplace add oodle-ai/agent-skills\n/plugin install oodle-agent-skills@oodle-ai\n```\n\nUpdate on demand with `/plugin marketplace update oodle-ai` (plugins otherwise auto-update in the background).\n\n**GitHub Copilot CLI** — also reads `.claude-plugin/marketplace.json` directly:\n\n```\ncopilot plugin marketplace add oodle-ai/agent-skills\ncopilot plugin install oodle-agent-skills@oodle-ai\n```\n\n**Factory.ai (Droid)** — reads `.factory-plugin/marketplace.json`:\n\n```\ndroid plugin marketplace add https://github.com/oodle-ai/agent-skills\ndroid plugin install oodle-agent-skills@oodle-ai\n```\n\n**Cursor** — reads `.cursor-plugin/marketplace.json`. Cursor onboards marketplaces through the GUI (no CLI add): Dashboard → **Plugins → Team Marketplaces → Add Marketplace → Import from Repo**, then paste `https://github.com/oodle-ai/agent-skills`. Requires the Cursor GitHub App on the repo; enable **Auto Refresh** to pull on push.\n\n> **Codex** ships no separate plugin format — it reuses the Claude Code layout. Install Codex skills with `npx skills` below.\n\n### Every agent: `npx skills`\n\nThe [`skills`](https://github.com/obra/skills) installer works across coding agents. Pick your `-a` agent slug (`-g` = global/user-level, `-y` = no prompts):\n\n| Agent | Command |\n|-------|---------|\n| Claude Code | `npx skills add oodle-ai/agent-skills -a claude-code -y -g` |\n| Cursor | `npx skills add oodle-ai/agent-skills -a cursor -y -g` |\n| Codex | `npx skills add oodle-ai/agent-skills -a codex -y -g` |\n| GitHub Copilot | `npx skills add oodle-ai/agent-skills -a copilot -y -g` |\n| Factory.ai (Droid) | `npx skills add oodle-ai/agent-skills -a factory -y -g` |\n| Gemini CLI | `npx skills add oodle-ai/agent-skills -a gemini-cli -y -g` |\n| Windsurf | `npx skills add oodle-ai/agent-skills -a windsurf -y -g` |\n\nInstall to several agents at once with a comma-separated list, e.g. `-a claude-code,cursor,codex,copilot,factory`, or `-a '*'` for all supported agents.\n\nNotes on what each agent picks up:\n\n- **Claude Code** installs skills under `~/.claude/skills/` and registers `.claude-plugin/plugin.json`. Prefer the native marketplace above when you want auto-updates.\n- **Cursor** installs into Cursor's rules directory using `.cursor-plugin/plugin.json` for plugin metadata.\n- **Codex / Copilot / Factory.ai** install into each agent's skills/rules directory; they consume the same `SKILL.md` documents.\n- **Gemini CLI** picks up the root `gemini-extension.json` manifest automatically.\n\n### Oodle CLI\n\nThe Oodle CLI ships its own skills installer that places every skill in the right location for the AI agent that's currently active in your shell.\n\n```bash\noodle skills install\n```\n\n### Manual install\n\nClone the repo and copy the `skills/` directory into your agent's skills location.\n\n```bash\ngit clone https://github.com/oodle-ai/agent-skills.git\ncp -r agent-skills/skills/* ~/.<agent>/skills/\n```\n\n## Quick Reference\n\n| Task | Command |\n|------|---------|\n| Configure auth | `oodle configure` |\n| List monitors | `oodle monitors list -o json` |\n| Get monitor | `oodle monitors get <id> -o json` |\n| Create monitor | `oodle monitors create -f monitor.json` |\n| Update monitor | `oodle monitors update <id> -f monitor.json` |\n| Delete monitor | `oodle monitors delete <id> --force` |\n| List dashboards | `oodle dashboards list -o json` |\n| Create dashboard | `oodle dashboards create -f dashboard.json` |\n| List folders | `oodle folders list -o json` |\n| List notifiers | `oodle notifiers list -o json` |\n| List notification policies | `oodle notification-policies list -o json` |\n| List muting rules | `oodle muting-rules list -o json` |\n| Search metrics | `oodle metrics list --match \"http_requests\"` |\n| List metric labels | `oodle metrics labels http_requests_total` |\n| List label values | `oodle metrics label-values http_requests_total service` |\n| Instant PromQL query | `oodle metrics query --query \"sum(up)\" -o json` |\n| Range PromQL query | `oodle metrics query-range --query \"up\" --start -1h --end now --step 60s -o json` |\n| List log index patterns | `oodle logs index-patterns -o json` |\n| Search logs | `oodle logs query -f query.ndjson -o json` |\n| Search traces | `oodle traces list --service api --from -1h` |\n| Get trace | `oodle traces get <trace-id> -o json` |\n| List log metrics | `oodle log-metrics list -o json` |\n| List synthetic monitors | `oodle synthetic-monitors list -o json` |\n| List drop rules | `oodle drop-rules list -o json` |\n| List integrations | `oodle integrations list -o json` |\n| Get integration setup spec | `oodle integrations get-setup-spec <type> -o json` |\n\n## Authoring Conventions\n\nEvery skill in this repo follows the same canonical structure:\n\n1. YAML frontmatter (name, description, metadata)\n2. Prerequisites\n3. Command Execution Order\n4. Quick Reference table\n5. Common Operations with `✅ CORRECT` / `❌ WRONG` examples\n6. Best Practices with prescriptive, testable rules\n7. Failure Handling table\n8. References\n\nRules in the skills are written prescriptively (`Run X` rather than `You can run X`) so that the agent treats them as hard constraints rather than suggestions.\n\n## Links\n\n- [Oodle CLI repo](https://github.com/oodle-ai/oodle-cli) — source of the `oodle` binary\n- [Oodle docs](https://docs.oodle.ai) — product documentation\n- [Issues](https://github.com/oodle-ai/agent-skills/issues) — report a problem or request a new skill\n\n## License\n\n[MIT](LICENSE)\n",
  "bytes": 8727,
  "sha": "ea9e8e0c8055e0cd3c4f45e76ea41ba836c5c43de08cbe9b49fc209d4f776c71",
  "repo_slug": "oodle-ai/agent-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_oodle_ai_agent_skills_33348b49/readme"
}