The maximalist: 18 agents, 50+ skills, and the advice to start with three

Run an academic research workflow with 18 agents and 50+ skills. Academics (econometrics / research papers) using Claude Code.

Share on 𝕏 ↗ Compare with… Read the source ↗

what to steal

Start with CLAUDE.md plus two or three skills; grow only from real friction.

Our takeaway from the source, not a quote.

Drawn from this card's sourced fields. Amber = older model named by the source. Rose = cut or replaced.

pedro-santanna-academic-claude-code.toml

# Run an academic research workflow with 18 agents and 50+ skills
# source: https://github.com/pedrohcgs/claude-code-my-workflow

[rig]
who     = "Academics (econometrics / research papers) using Claude Code"
author  = "Pedro H. C. Sant'Anna"
date    = "not stated"

harness = ["Claude Code"]
models  = []

[keep]
  - 18 agents
  - 50+ skills
  - 32 rules
  - 7 hooks
  - MIT-licensed repo

thing	status	why (per source)
Starting point	kept small	Advises: "begin with CLAUDE.md and two or three skills"

Source: https://github.com/pedrohcgs/claude-code-my-workflow · https://psantanna.com/claude-code-my-workflow/ · checked 2026-10-14

A dedicated repo where research agents get the open internet

steal → Give risky, network-hungry research its own repo instead of loosening your main one.

CodexClaude Code Network lockdown

by Simon Willison

~/rigs/mrclaw207-openclaw-247.tomlundated

One OpenClaw agent 24/7 plus one research subagent

steal → Start with one main agent and one helper, not a swarm.

OpenClawClaude Code claude-sonnet-4-202505 Many agents up front

by mrclaw207 (DEV)

~/rigs/merlin-claude-manager-codex-….toml𝕏 · 2026-10-14

Claude manages, Codex grinds

steal → Use one model to plan architecture and another to churn through build steps.

Claude CodeCodex Claude Fable as manageGPT-6 Astra on Codex

by @Merlin412188240