/storm:research
Phase 1 of the STORM pipeline. Discovers research personas, runs a simulated multi-turn conversation per persona (each grounded in retrieval), and produces the information table that downstream phases consume.
CRITICAL: Load Engine First
Load storm-engine via the Skill tool. Its "Persona Discovery", "Simulated Conversation", "Retrieval", and "Citation Hygiene" sections govern this phase.
Completion Contract
This phase is complete iff research/sources.json exists and has ≥1 entry. If --force is not set and the artifact exists, skip and exit early.
Procedure
- Resolve output dir (see engine). Ensure
research/ subdir exists.
- Persona discovery — search the web for the topic + 2-3 related concepts; fetch reference pages and extract their section headings. Use those real structures to propose
--max-perspective personas (default 3), each with a distinct question category, plus one "Basic fact writer". Write research/personas.json.
- Retrieval probe — via
ToolSearch, look for exa-mcp-server search tools (code-search, research-paper-search, company-search, personal-site-search, financial-report-search, x-search). Record which are available in run-config.json as retriever. If none and not --docs, fall back to WebSearch/WebFetch.
- Parallel simulated conversations — launch one
storm-researcher subagent per persona in a single message (parallel). Each subagent:
- Receives: topic, persona definition, retrieval instructions,
max_turns.
- Runs the WikiWriter↔TopicExpert dialogue: writer asks a question, expert does
question_to_query, retrieves, answers with source attribution.
- Ends when writer says "Thank you so much for your help!" or
max_turns reached.
- Returns: a JSON array of
{question, queries, snippets, answer, cited_sources}.
- Launch all persona subagents in a single message (parallel) — do not serialize.
- Merge — collect all conversation records into
research/conversations.jsonl (one JSON object per line). Deduplicate cited sources by URL into research/sources.json, assigning sequential ids. Strip any inline [n] from snippets before storing (citation hygiene).
- Verify — assert
sources.json has ≥1 entry and every persona produced ≥1 turn. If a persona produced nothing, note it but do not fail the whole phase.
- Update
run-config.json: phases.research = "completed", retriever, source count.
Concurrency Note
If the user reports rate-limit errors during this phase, reduce concurrency by running personas in smaller batches (e.g. 2 at a time) rather than lowering max_turns — mirroring upstream's max_thread_num guidance.
Output
Report: number of personas, total conversation turns, number of deduplicated sources, and the path to research/sources.json.
1---2name: storm-research3description: Run STORM phase 1 — multi-perspective research. This skill should be used when the user asks to "research a topic for an article", "find sources on X from multiple perspectives", "do storm research on X", or invokes /storm:research. Discovers personas, runs simulated Q&A per persona in parallel, and produces an information table + deduplicated sources.4---56# /storm:research78Phase 1 of the STORM pipeline. Discovers research personas, runs a simulated multi-turn conversation per persona (each grounded in retrieval), and produces the information table that downstream phases consume.910## CRITICAL: Load Engine First1112Load `storm-engine` via the Skill tool. Its "Persona Discovery", "Simulated Conversation", "Retrieval", and "Citation Hygiene" sections govern this phase.1314## Completion Contract1516This phase is complete iff `research/sources.json` exists and has ≥1 entry. If `--force` is not set and the artifact exists, skip and exit early.1718## Procedure19201. Resolve output dir (see engine). Ensure `research/` subdir exists.212. **Persona discovery** — search the web for the topic + 2-3 related concepts; fetch reference pages and extract their section headings. Use those real structures to propose `--max-perspective` personas (default 3), each with a distinct question category, plus one "Basic fact writer". Write `research/personas.json`.223. **Retrieval probe** — via `ToolSearch`, look for exa-mcp-server search tools (`code-search`, `research-paper-search`, `company-search`, `personal-site-search`, `financial-report-search`, `x-search`). Record which are available in `run-config.json` as `retriever`. If none and not `--docs`, fall back to `WebSearch`/`WebFetch`.234. **Parallel simulated conversations** — launch one `storm-researcher` subagent per persona in a single message (parallel). Each subagent:24 - Receives: topic, persona definition, retrieval instructions, `max_turns`.25 - Runs the WikiWriter↔TopicExpert dialogue: writer asks a question, expert does `question_to_query`, retrieves, answers with source attribution.26 - Ends when writer says "Thank you so much for your help!" or `max_turns` reached.27 - Returns: a JSON array of `{question, queries, snippets, answer, cited_sources}`.28 - Launch all persona subagents in a single message (parallel) — do not serialize.295. **Merge** — collect all conversation records into `research/conversations.jsonl` (one JSON object per line). Deduplicate cited sources by URL into `research/sources.json`, assigning sequential `id`s. Strip any inline `[n]` from snippets before storing (citation hygiene).306. **Verify** — assert `sources.json` has ≥1 entry and every persona produced ≥1 turn. If a persona produced nothing, note it but do not fail the whole phase.317. Update `run-config.json`: `phases.research = "completed"`, `retriever`, source count.3233## Concurrency Note3435If the user reports rate-limit errors during this phase, reduce concurrency by running personas in smaller batches (e.g. 2 at a time) rather than lowering `max_turns` — mirroring upstream's `max_thread_num` guidance.3637## Output3839Report: number of personas, total conversation turns, number of deduplicated sources, and the path to `research/sources.json`.
FradSer