vmeyer/research-toolkit
Multi-IDE research-and-summarize agent workflow template for Claude Code, Gemini, Copilot, and Windsurf
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit vmeyer/research-toolkit/.claude ./rig-research-toolkit # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit vmeyer/research-toolkit/.claude/skills/research-and-summarize .claude/skills/research-and-summarize $ npx degit vmeyer/research-toolkit/.claude/skills/research-dashboard .claude/skills/research-dashboard
$ curl -fsSL --create-dirs -o .claude/agents/brief-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/brief-1.md $ curl -fsSL --create-dirs -o .claude/agents/decision-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/decision-1.md $ curl -fsSL --create-dirs -o .claude/agents/detailed-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/detailed-1.md $ curl -fsSL --create-dirs -o .claude/agents/final-critic-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/final-critic-1.md $ curl -fsSL --create-dirs -o .claude/agents/html-report-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/html-report-1.md $ curl -fsSL --create-dirs -o .claude/agents/intake-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/intake-1.md $ curl -fsSL --create-dirs -o .claude/agents/keypoints-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/keypoints-1.md $ curl -fsSL --create-dirs -o .claude/agents/researcher-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/researcher-1.md $ curl -fsSL --create-dirs -o .claude/agents/verifier-1.md https://raw.githubusercontent.com/vmeyer/research-toolkit/main/.claude/agents/verifier-1.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (2)
Subagents (9)
| brief-1 model: sonnet | Write brief summary and save to file |
| decision-1 model: opus | Produce a decision document (ADR style) from verified research — prioritized recommendation, alternatives, decision criteria, risks, and PoC/acceptance criteria. |
| detailed-1 model: sonnet | Write detailed report and save to file |
| final-critic-1 model: opus | Read-only final critic for high-assurance research runs. Checks the synthesized draft against the evidence contract, claims, and sources without editing anything. |
| html-report-1 model: opus | Produce styled HTML report from template |
| intake-1 model: opus | Clarify research topic iteratively and produce research brief with sub-briefs |
| keypoints-1 model: sonnet | Extract key points for skill creation and save to file |
| researcher-1 model: sonnet | Research a sub-brief using web search with triangulation strategy |
| verifier-1 model: opus | Synthesize parallel research results, analyze themes, and verify quality |
Slash commands (2)
/research-and-summarize/research-dashboard
Similar rigs
JasonCodeMaker/Agentic-Research-Control-Panel
Local control plane for governed AutoResearch in real repos: human-approved scope, live experiment tracking, evidence-backed results, and project memory for Claude Code and Codex.
Skill Collector 1.8k tok ·
DevelopersGlobal/ai-agent-skills
AI agent skills for production grade applications
Skill Collector 2.0k tok ·
rtk-ai/rtk
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
Orchestrator 2.6k tok ·
Yeachan-Heo/oh-my-claudecode
Teams-first Multi-agent orchestration for Claude Code
Orchestrator 14.5k tok ·