# search_specialist

The search_specialist Claude Code subagent turns an information need into focused queries and assesses retrieved sources. It returns a compact evidence set in JSON for downstream work.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 776 repo stars
- Author: AgenticOrgLab
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/AgenticOrgLab/ExecOrchestra/tree/HEAD/execorchestra/Pool/resources/agents/search_specialist.md
- Updated: 2026-11-12
- Security check: automated, 2026-11-17 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/agenticorglab-search-specialist

## What it does

search_specialist builds queries from a search objective, assesses retrieved material for relevance and authority, then returns a compact set of supported facts. It keeps source claims distinct from relevance assessments and records gaps and duplicate sources.

Its inputs are the search objective, including scope and freshness requirements, plus context such as keywords, known sources, earlier results, and exclusions. The JSON output contains queries, source details and relevance reasons, coverage gaps, and notes on overlapping results. It reports a source as retrieved only when that source’s content is supplied.

Suggested tools are routing options, not a required allowlist; usable resources depend on the resource application plan. For Claude Code, place the agent file at `.claude/agents/search_specialist.md`.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/search_specialist.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/search_specialist.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/search_specialist.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/AgenticOrgLab/ExecOrchestra/tree/HEAD/execorchestra/Pool/resources/agents/search_specialist.md

## When to use it
- turn a search objective and its constraints into focused queries
- separate source-supported facts from relevance judgments
- identify unresolved search areas and overlapping sources

## Example prompts
- Find primary sources on this topic and return supported facts with source details as JSON.
- Build search queries for this question, account for freshness, and exclude the results already listed.
- Assess these sources for relevance, then report gaps and duplicates.

## Similar
- [web-search](https://affpapa.org/en/nevibecode/skills/claude-office-skills-web-search.md) — skill · 5000 installs
- [exa-search](https://affpapa.org/en/nevibecode/skills/affaan-m-exa-search.md) — skill · 6300 installs
- [tavily-search](https://affpapa.org/en/nevibecode/skills/framix-team-tavily-search.md) — skill · 843 installs
- [paper-search](https://affpapa.org/en/nevibecode/plugins/paper-search.md) — plugin · ★ 67 repo stars
- [perplexity-search](https://affpapa.org/en/nevibecode/skills/perplexity-search.md) — skill · 536 installs
- [deep-research](https://affpapa.org/en/nevibecode/skills/deep-research.md) — skill · 83 installs

From the same set: [agenticorglab/execorchestra](https://affpapa.org/en/nevibecode/by/agenticorglab/execorchestra.md) — 24 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
