How aammartufail/top_five_ai_agents_for_research works
This repository is a starter kit for five independent AI research agents, plus guides and learning materials. The main runtime workflow spans discovery, systematic review, experimentation, manuscript preparation and PhD mentoring; the agents are separate CLIs rather than a connected pipeline. The sampled entrypoints support the workflow and orchestration edges shown here; unsampled agent internals are represented only where the README or entrypoint establishes their role.
Main entry points: agent_03_research_experiment/main.py Experiment CLI · agent_04_academic_writing/main.py Writing CLI · llm.py LLM Client · agent_01_research_discovery/main.py Discovery CLI
Components 29
Research Discovery
- Discovery CLI
02_agents_code/agent_01_research_discovery/main.py - Paper Search
02_agents_code/agent_01_research_discovery/tools.py - Trends & Gaps
02_agents_code/agent_01_research_discovery/agents.py - Ideas & Proposal
02_agents_code/agent_01_research_discovery/agents.py
Literature Review
- Review CLI
02_agents_code/agent_02_literature_review/main.py - Review Search
02_agents_code/agent_02_literature_review/tools.py - Screen & Extract
02_agents_code/agent_02_literature_review/agents.py - PRISMA Counts
02_agents_code/agent_02_literature_review/prisma.py
Research Experiment
- Experiment CLI
02_agents_code/agent_03_research_experiment/main.py - Data Input
02_agents_code/agent_03_research_experiment/data.py - Statistical & ML
02_agents_code/agent_03_research_experiment/analysis.py - Design & Reporting
02_agents_code/agent_03_research_experiment/agents.py - Notebook Output
02_agents_code/agent_03_research_experiment/notebook.py
Writing & Mentoring
- Writing CLI
02_agents_code/agent_04_academic_writing/main.py - Manuscript Drafting
02_agents_code/agent_04_academic_writing/agents.py - Journal Matching
02_agents_code/agent_04_academic_writing/agents.py - Reviewer Simulation
02_agents_code/agent_04_academic_writing/agents.py - Revision & Rebuttal
02_agents_code/agent_04_academic_writing/agents.py - Supervisor CLI
02_agents_code/agent_05_phd_supervisor/main.py - Mentoring Tasks
02_agents_code/agent_05_phd_supervisor/agents.py - Student Memory
02_agents_code/agent_05_phd_supervisor/memory.py
Shared Agent Layer
- LLM Client
02_agents_code/agent_01_research_discovery/llm.py - LLM Client
02_agents_code/agent_02_literature_review/llm.py - LLM Client
02_agents_code/agent_03_research_experiment/llm.py - LLM Client
02_agents_code/agent_04_academic_writing/llm.py - LLM Client
02_agents_code/agent_05_phd_supervisor/llm.py
Other components
- Researcher
- Anthropic API
- Research APIs
Connections 37
- Researcher → Discovery CLI: submits topic
- Discovery CLI → Paper Search: searches papers
- Paper Search → Research APIs: queries APIs
- Discovery CLI → Trends & Gaps: analyzes literature
- Discovery CLI → Ideas & Proposal: generates outputs
- Trends & Gaps → LLM Client: requests analysis
- Ideas & Proposal → LLM Client: generates text
- LLM Client → Anthropic API: calls model
- Researcher → Review CLI: submits question
- Review CLI → Review Search: searches records
- Review Search → Research APIs: queries APIs
- Review CLI → Screen & Extract: screens and writes
- Review CLI → PRISMA Counts: builds counts
- Screen & Extract → LLM Client: requests analysis
- LLM Client → Anthropic API: calls model
- Researcher → Experiment CLI: describes study
- Experiment CLI → Data Input: loads data
- Experiment CLI → Statistical & ML: runs analyses
- Experiment CLI → Design & Reporting: plans and reports
- Experiment CLI → Notebook Output: builds notebook
- Design & Reporting → LLM Client: requests analysis
- LLM Client → Anthropic API: calls model
- Researcher → Writing CLI: provides notes
- Writing CLI → Manuscript Drafting: drafts manuscript
- Writing CLI → Journal Matching: matches journals
- Writing CLI → Reviewer Simulation: simulates review
- Writing CLI → Revision & Rebuttal: revises manuscript
- Manuscript Drafting → LLM Client: requests drafting
- Journal Matching → LLM Client: requests matching
- Reviewer Simulation → LLM Client: requests review
- Revision & Rebuttal → LLM Client: requests revision
- LLM Client → Anthropic API: calls model
- Researcher → Supervisor CLI: submits updates
- Supervisor CLI → Student Memory: reads and writes
- Supervisor CLI → Mentoring Tasks: runs mentoring tasks
- Mentoring Tasks → LLM Client: requests guidance
- LLM Client → Anthropic API: calls model