deep-researchlisted

Expanded, multi-angle web research - an upgrade over a single WebSearch. Decomposes a topic into angles, runs parallel researcher subagents (one per angle, source-cited and dated), cross-validates their findings for contradictions and source authority, then synthesizes one confidence-rated, source-cited brief. Provider- and project-agnostic; writes nothing outside its output unless asked. Use for "deep research", "research X thoroughly", "multi-angle", "what's the current state of X", "is X still true", any request for an answer that is "verified, not guessed" (--inline covers quick verified lookups), or any question where a single search is too shallow and accuracy matters.
anton-winter-arch/dotagents · ★ 1 · AI & Automation · score 74

Install: claude install-skill anton-winter-arch/dotagents

# deep-research A disciplined upgrade over a one-shot `WebSearch`: parallel angle coverage + adversarial cross-checking + dated, source-cited synthesis. It **gathers and verifies facts**; it does not give opinions, and it touches no project files unless you pass `--save`. ``` 1. PLAN (main agent) decompose the topic into 2-5 angles (5-7 for --deep) 2. RESEARCH (N parallel agents) one researcher per angle: web search, dated, source-cited 3. VALIDATE (1 agent) cross-check claims, flag contradictions, rate authority 4. SYNTHESIZE (main agent) one confidence-rated, source-cited brief ``` ## When to use - Explicit `/deep-research <topic>`, or "research X thoroughly / from multiple angles". - Currency-sensitive questions ("current state of", "is X still true", "latest on"). - Before building on a fast-moving fact (a model/version/pricing/API claim) where being wrong is expensive. ## When NOT to use - A one-line fact lookup → use raw `WebSearch`. - The answer is already in the repo/workspace → read the source. - The user wants a recommendation/opinion → research surfaces facts; decide separately. ## Modes | Invocation | Pipeline | When | |---|---|---| | `/deep-research <topic>` | Full: PLAN → researchers (max 3 concurrent) → VALIDATE → SYNTHESIZE | Default. Substantive topic. | | `--inline <topic>` | Single agent: 3-5 searches + inline validation, no subagent spawns | Quick check that still needs verification. | | `--deep <topic>` | Full wit