home / skills / plurigrid / asi / depth-search

depth-search skill

/skills/depth-search

This skill performs deep multi-source research by querying local ~/.topos, academic MCPs, and Exa semantic search to synthesize comprehensive results.

npx playbooks add skill plurigrid/asi --skill depth-search

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
3.7 KB
---
name: depth-search
description: Deep multi-source research combining academic MCPs (arxiv, semantic-scholar, paper-search, deepwiki), Exa semantic search, and local ~/.topos knowledge base. Use for comprehensive research requiring multiple sources. NEVER fall back to web_search - ask user for help instead.
---

# Depth Search

Comprehensive multi-source research skill. Searches across academic databases, semantic web search, and local knowledge before asking the user for help.

## Search Order

Execute searches in this order, using parallel subagents where possible:

### 1. Local Knowledge Base (~/.topos)
Search `~/.topos` directory first for existing research, notes, and cached data:
- Use `glob` and `Grep` to find relevant files
- Check `.md`, `.org`, `.jl`, `.py`, `.json` files
- Look in subdirectories: `skills/`, `archived/`, `Gay.jl/`, etc.

### 2. Academic MCPs (parallel)
Launch parallel subagents to search all 4 academic sources:

| MCP | Tools | Best For |
|-----|-------|----------|
| **arxiv** | `search_papers`, `get_paper`, `download_paper` | Preprints, CS/physics/math papers |
| **semantic-scholar** | `paper_relevance_search`, `paper_details`, `paper_citations` | Citation analysis, author profiles |
| **paper-search** | `search_arxiv`, `search_pubmed`, `search_biorxiv`, etc. | Multi-source aggregation |
| **deepwiki** | `read_wiki_structure`, `read_wiki_contents`, `ask_question` | GitHub repo documentation |

### 3. Exa Semantic Search
Use Exa MCP for high-quality web search:
- `web_search_exa` - Semantic web search
- `crawling_exa` - Extract web content
- `company_research_exa` - Company research
- `deep_researcher_start` / `deep_researcher_check` - Deep research tasks

### 4. Ask User for Help
If all sources fail to find what's needed:
- **DO NOT fall back to `web_search`** - it's basic keyword matching only
- Instead, ask the user:
  - "I couldn't find [X] in academic databases, Exa, or local files. Can you provide a link, paper title, or more context?"
  - Suggest specific sources they might check manually
  - Offer to try different search terms

## Critical Rules

1. **NEVER use `web_search` as a fallback** - it's not equivalent to Exa
2. **NEVER use `web_search` in Task subagents** - use Exa tools instead
3. **Always search local ~/.topos first** - may have cached/annotated versions
4. **Use parallel subagents** for academic MCPs to maximize speed
5. **Ask user for help** rather than guessing or using inferior search

## Example Workflow

```
User: "Find papers on world models for LLMs"

1. Search ~/.topos for existing notes/papers
2. Launch 4 parallel Task subagents:
   - arxiv: search_papers("world models LLM")
   - semantic-scholar: paper_relevance_search("world models language models")
   - paper-search: search across all sources
   - deepwiki: check relevant GitHub repos
3. If needed, use Exa: web_search_exa("world models LLM research")
4. Synthesize results from all sources
5. If still not found: ask user for clarification
```

## Parallel Subagent Template

When searching academic sources, use this pattern:

```
Launch 4 parallel Task subagents:
- Task 1: Use arxiv MCP to search for [query]
- Task 2: Use semantic-scholar MCP to search for [query]  
- Task 3: Use paper-search MCP to search for [query]
- Task 4: Use deepwiki MCP to find related repos/docs
```

## What NOT To Do

❌ `web_search` as fallback when Exa fails  
❌ Single-source search when multiple are available  
❌ Skipping local ~/.topos search  
❌ Guessing answers without exhausting sources  
❌ Sequential searches when parallel is possible  

## What TO Do

✅ Search ~/.topos first for cached knowledge  
✅ Parallel subagents for academic MCPs  
✅ Exa for semantic web search  
✅ Ask user when sources are exhausted  
✅ Synthesize results from multiple sources  

Overview

This skill performs deep multi-source research by combining a prioritized local knowledge base with multiple academic MCPs and a high-quality semantic web search. It is designed for comprehensive literature and repo discovery, synthesizing results from local files, arXiv, Semantic Scholar, paper-search, deepwiki, and Exa. The workflow enforces parallel searches across academic sources and never falls back to basic web_search; instead it asks the user for guidance when needed.

How this skill works

First, the skill inspects the local ~/.topos directory for cached notes, code, and prior research using glob and grep across common file types (.md, .org, .jl, .py, .json). Next it launches parallel subagents to query arXiv, Semantic Scholar, paper-search, and deepwiki for preprints, citations, aggregated sources, and repository docs. After academic aggregation it runs Exa semantic search tools to extract web content and company/research signals. If none of these sources yield sufficient evidence, it prompts the user for links, titles, or clarification rather than using inferior web_search fallbacks.

When to use it

  • Comprehensive literature reviews requiring coverage across preprints, citations, and docs
  • Investigations where local cached notes or prior experiments may exist in ~/.topos
  • Fast, parallelized searches to reduce latency across multiple academic sources
  • Synthesis tasks that must combine paper content, citation context, and repo docs
  • Situations where guessing is unacceptable and user clarification is preferable

Best practices

  • Always run the local ~/.topos search first to pick up cached annotations and experiments
  • Launch academic MCP queries in parallel to maximize throughput and avoid sequential bottlenecks
  • Use Exa MCP tools for semantic web crawling and company/research extraction—never substitute web_search
  • If results are incomplete, ask the user for a link, exact title, DOI, or additional keywords
  • Record and cache synthesized findings back into ~/.topos to speed future research

Example use cases

  • Find recent preprints and citation networks on world models for large language models
  • Locate implementation details and README docs for a chemistry automation repo referenced in notes
  • Combine local lab notebooks with published methods to prepare a reproducible protocol summary
  • Perform company and product research using Exa after aggregating academic evidence
  • Search for conflicting results across papers, then request the user to resolve ambiguous references

FAQ

What happens if Exa and academic MCPs return no useful results?

I will not use basic web_search; instead I will ask you for a link, paper title, DOI, or more context to continue the search.

Why search ~/.topos first?

Local files may contain cached searches, annotated papers, or experiment notes that save time and avoid duplicated work.