Literature Survey
Autonomous Literature Survey Engine — select the right paradigm, execute with full autonomy, deliver structured output.
Design Philosophy
Strategy Book — autonomous execution, minimal dialogue.
After brief intake (if needed), CC selects a strategy and executes autonomously. The user receives a structured survey output, not a conversation. CC has full autonomy over tactic sequencing, iteration, budget allocation, and saturation judgment. The only hard constraints are the budget table (quantitative floors, ±10% flexibility) and quality gates from imported base SOPs.
Routing
Determine which strategy to invoke based on user context:
| Signal |
Strategy |
| New field, "what exists", landscape, orientation |
→ scoping-survey |
| Exhaustive coverage, "all methods", PRISMA, rigorous |
→ systematic-survey |
| Specific sub-problem, precise understanding, technical depth |
→ deep-survey |
| Argument building, position paper, theoretical framework |
→ narrative-review |
| Has seed papers, citation tracing, lineage, evolution |
→ snowball |
If ambiguous: invoke intake SOP (1-3 questions max) to clarify intent, then route.
Available MCP Tools
CC decides which tools to use and when. These are available throughout execution:
| MCP Server |
Tools |
| brave-search |
brave_web_search, brave_news_search, brave_llm_context |
| apify |
rag-web-browser, google-scholar-scraper |
| alphaxiv |
discover_papers, get_paper_content, answer_pdf_queries, read_files_from_github_repository |
| semantic-scholar |
ss_paper, ss_paper_batch, ss_references, ss_citations, ss_recommendations, ss_relevance_search, ss_author, ss_author_papers |
Layer Structure
Entry (this file) → routes to 1 of 5 strategies
Strategy → budget + available tactics/SOPs + execution guidance
Tactic → multi-step orchestration pattern
SOP → single focused operation (dialogue / subagent / import)
Each layer calls only the layer directly below it.
1---2name: literature-survey-33description: Autonomous Literature Survey Engine with 5 research paradigms (scoping, systematic, deep, narrative, snowball). Use this skill whenever a user needs to survey academic literature, conduct a literature review, map a research field, trace citation lineages, or build evidence-based arguments from papers — regardless of how specific or vague their request is. Also use when the user mentions systematic review, PRISMA, snowball sampling, scoping review, or narrative review.4---56# Literature Survey78Autonomous Literature Survey Engine — select the right paradigm, execute with full autonomy, deliver structured output.910## Design Philosophy1112**Strategy Book — autonomous execution, minimal dialogue.**1314After brief intake (if needed), CC selects a strategy and executes autonomously. The user receives a structured survey output, not a conversation. CC has full autonomy over tactic sequencing, iteration, budget allocation, and saturation judgment. The only hard constraints are the budget table (quantitative floors, ±10% flexibility) and quality gates from imported base SOPs.1516## Routing1718Determine which strategy to invoke based on user context:1920| Signal | Strategy |21|--------|----------|22| New field, "what exists", landscape, orientation | → scoping-survey |23| Exhaustive coverage, "all methods", PRISMA, rigorous | → systematic-survey |24| Specific sub-problem, precise understanding, technical depth | → deep-survey |25| Argument building, position paper, theoretical framework | → narrative-review |26| Has seed papers, citation tracing, lineage, evolution | → snowball |2728**If ambiguous**: invoke `intake` SOP (1-3 questions max) to clarify intent, then route.2930## Available MCP Tools3132CC decides which tools to use and when. These are available throughout execution:3334| MCP Server | Tools |35|------------|-------|36| brave-search | brave_web_search, brave_news_search, brave_llm_context |37| apify | rag-web-browser, google-scholar-scraper |38| alphaxiv | discover_papers, get_paper_content, answer_pdf_queries, read_files_from_github_repository |39| semantic-scholar | ss_paper, ss_paper_batch, ss_references, ss_citations, ss_recommendations, ss_relevance_search, ss_author, ss_author_papers |4041## Layer Structure4243```44Entry (this file) → routes to 1 of 5 strategies45 Strategy → budget + available tactics/SOPs + execution guidance46 Tactic → multi-step orchestration pattern47 SOP → single focused operation (dialogue / subagent / import)48```4950Each layer calls only the layer directly below it.
yogsoth-ai