{"id":5648,"plugin_id":"plugin_asdk_app_69f271663a288191ac98f46bed7cb032","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T22:45:46.280Z","digest":"7e3ec9eca32917543c37689e8c499f1b61527aa5312b86c48b61f18b850ac780","against":null,"payload":{"name":"tavily-research","description":"Conduct comprehensive AI-powered research with citations via the Tavily CLI. Use this skill when the user wants deep research, a detailed report, a comparison, market analysis, literature review, or says \"research\", \"investigate\", \"analyze in depth\", \"compare X vs Y\", \"what does the market look like for\", or needs multi-source synthesis with explicit citations. Returns a structured report grounded in web sources. Takes 30-120 seconds. For quick fact-finding, use tavily-search instead.\n","included_files":[],"skill_md_contents":"---\nname: tavily-research\ndescription: |\n  Conduct comprehensive AI-powered research with citations via the Tavily CLI. Use this skill when the user wants deep research, a detailed report, a comparison, market analysis, literature review, or says \"research\", \"investigate\", \"analyze in depth\", \"compare X vs Y\", \"what does the market look like for\", or needs multi-source synthesis with explicit citations. Returns a structured report grounded in web sources. Takes 30-120 seconds. For quick fact-finding, use tavily-search instead.\nallowed-tools: Bash(tvly *)\n---\n\n# tavily research\n\nAI-powered deep research that gathers sources, analyzes them, and produces a cited report. Takes 30-120 seconds.\n\n## Before running\n\nResearch requires authentication. Run the requested command directly when\n`tvly` is already authenticated; do not add a status check to every invocation.\n\nIf `tvly` is missing, follow the [tavily-cli setup](../tavily-cli/SKILL.md#setup).\nIf an installed CLI reports an authentication error, use `tvly login` for\nauthentication only, or `tvly init --skip-skills` when guided verification is\nalso useful. Browser-based OAuth is preferred when an interactive user can\ncomplete it. `--no-browser` prints the sign-in link instead of opening it, but\nstill waits for a localhost callback. In an unattended agent or CI environment,\nleave authentication to the user or use a securely provided `TAVILY_API_KEY`.\nDo not start a second login immediately after guided setup has completed.\n\n## When to use\n\n- You need comprehensive, multi-source analysis\n- The user wants a comparison, market report, or literature review\n- Quick searches aren't enough — you need synthesis with citations\n- Step 5 in the [workflow](../tavily-cli/SKILL.md): search → extract → map → crawl → **research**\n\n## Quick start\n\n```bash\n# Basic research (waits for completion)\ntvly research \"competitive landscape of AI code assistants\"\n\n# Pro model for comprehensive analysis\ntvly research \"electric vehicle market analysis\" --model pro\n\n# Stream results in real-time\ntvly research \"AI agent frameworks comparison\" --stream\n\n# Save report to file\ntvly research \"fintech trends 2025\" --model pro -o fintech-report.json\n\n# JSON output for agents\ntvly research \"quantum computing breakthroughs\" --json\n```\n\n## Options\n\n| Option | Description |\n|--------|-------------|\n| `--model` | `mini`, `pro`, or `auto` (default) |\n| `--stream` | Stream results in real-time |\n| `--no-wait` | Return request_id immediately (async) |\n| `--output-schema` | Path to JSON schema for structured output |\n| `--citation-format` | `numbered`, `mla`, `apa`, `chicago` |\n| `--poll-interval` | Seconds between checks (default: 10) |\n| `--timeout` | Max wait seconds (default: 600) |\n| `-o, --output` | Save the JSON response to a file |\n| `--json` | Structured JSON output |\n\n## Model selection\n\n| Model | Use for | Speed |\n|-------|---------|-------|\n| `mini` | Single-topic, targeted research | ~30s |\n| `pro` | Comprehensive multi-angle analysis | ~60-120s |\n| `auto` | API chooses based on complexity | Varies |\n\n**Rule of thumb:** \"What does X do?\" → mini. \"X vs Y vs Z\" or \"best way to...\" → pro.\n\n## Async workflow\n\nFor long-running research, you can start and poll separately:\n\n```bash\n# Start without waiting\ntvly research \"topic\" --no-wait --json    # returns request_id\n\n# Check status\ntvly research status <request_id> --json\n\n# Wait for completion\ntvly research poll <request_id> --json -o result.json\n```\n\n## Tips\n\n- **Research takes 30-120 seconds** — use `--stream` to see progress in real-time.\n- **Use `--model pro`** for complex comparisons or multi-faceted topics.\n- **Use `--output-schema`** to get structured JSON output matching a custom schema.\n- **For quick facts**, use `tvly search` instead — research is for deep synthesis.\n- Read from stdin: `echo \"query\" | tvly research - --json`\n\n## See also\n\n- [tavily-search](../tavily-search/SKILL.md) — quick web search for simple lookups\n- [tavily-crawl](../tavily-crawl/SKILL.md) — bulk extract from a site for your own analysis\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}