/SKILL

/storm:generate

End-to-end STORM article generation. Runs all four phases in sequence, skipping any phase whose artifact already exists (unless --force).

CRITICAL: Load Engine First

Load the storm-engine skill via the Skill tool before doing anything else. It defines the artifact layout, stage-gating contract, citation hygiene, and retrieval fallback that this orchestration depends on. Do not improvise these.

Arguments

• <topic> (required) — the subject of the article.
• --max-perspective N (default 3) — number of personas to discover.
• --max-turns N (default 3) — max Q&A turns per persona.
• --output-dir PATH — explicit output location. Mutually exclusive with --save.
• --save — persist to docs/storm/<slug>/ instead of a temp dir.
• --docs DIR — ground on local documents in addition to (or, with --docs-only, instead of) the web.
• --force — re-run all phases even if artifacts exist.
• --retriever mcp|web|local — override retrieval source (default: mcp with fallback).

Procedure

1. Parse arguments; if no topic, invoke AskUserQuestion to request one.
2. Derive <slug> and resolve <output_dir> per the engine contract. Create the directory.
3. Write run-config.json with a snapshot of all parameters. Stamp started_at using date -u +%Y-%m-%dT%H:%M:%SZ (the only place a timestamp is generated).
4. For each phase in order — research, outline, write, polish: a. Check the phase's completion artifact (per engine contract). b. If complete and not --force: mark phases.<name>: "skipped" in run-config.json, log "Skipping (artifact present)". c. Otherwise: invoke the corresponding /storm:<phase> skill via the Skill tool, passing the resolved output dir and run parameters. After it returns, verify its artifact and mark phases.<name>: "completed" (or "failed" with the error).
5. After all phases, read article-polished.md and print a summary: topic, word count, number of sources cited, number of sections, and the absolute path to the final article.
6. If the run used a temporary directory, surface its absolute path prominently so the user can rescue artifacts.

Concurrency

research and write phases use Task subagents internally (see their own SKILL.md files). This orchestration skill runs phases strictly sequentially — never parallelize across phases, because each phase depends on the previous phase's artifacts.

Failure Handling

If a phase fails, stop and report. Do not proceed to the next phase. Update run-config.json with the failure reason. A re-invocation (without --force) will resume from the failed phase since completed prior phases will be skipped.

Output

The final message to the user must include:

• The absolute path to article-polished.md.
• Word count and section count.
• Number of cited sources.
• The output directory (especially if temporary).
• A one-line note that re-running without --force resumes from the first incomplete phase.