Skill

synapse

Orchestrates research projects, experiments, literature reviews, and compute resources via MCP tools for AI agents and humans. Useful for research workflows in Claude Code.

Anthropic

ai-ml

npx claudepluginhub vincentwei1021/synapse

Tool Access

This skill uses the workspace's default tool permissions.

Preview

Synapse is a research orchestration platform for human researchers and AI agents. Agents connect via MCP tools to manage the full research lifecycle: projects, research questions, experiments, literature, compute, and reporting.

Supporting Assets

package.jsonreferences/00-common-tools.mdreferences/01-setup.mdreferences/02-research-workflow.mdreferences/03-experiment-workflow.mdreferences/04-autonomous-loop.mdreferences/05-session-sub-agent.mdreferences/06-claude-code-agent-teams.md

SKILL.md

Similar Skills

Research Workflow — Full Pipeline

Orchestrates full research pipeline from Brainstorming to Reporting via Planning, Implementation, Testing & Visualization phases with user checkpoints. Configurable for physics, AI/ML, statistics, math domains, depth, and agent personas.

magi-researchers

research-manager

Orchestrates multi-step research projects by scaffolding plans, generating atomic tasks with dependencies, and persisting state via Tasks or files. Use for new research or complex workflows.

co-researcher

ARIS — Autonomous Research In Sleep

Automates ML research workflows with idea discovery, GPU-ready experiment code gen and W&B tracking, LaTeX paper writing, rebuttals, and cross-model review loops via Claude Code.

aradotso-trending-skills-37

Stats

Parent Repo Stars10

Parent Repo Forks1

Last CommitApr 10, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

Synapse Skill

Skill Files

File	Description
SKILL.md (this file)	Overview and getting started
references/00-common-tools.md	All MCP tools by category
references/01-setup.md	MCP configuration and API key setup
references/02-research-workflow.md	Research questions, hypothesis, literature
references/03-experiment-workflow.md	Experiment execution and compute
references/04-autonomous-loop.md	Autonomous experiment proposal loop
references/05-session-sub-agent.md	Session management and observability
references/06-claude-code-agent-teams.md	Claude Code Agent Teams integration

Core Workflow

ResearchProject --> ResearchQuestion --> Experiment --> Report
       ^                  ^                  ^            ^
     Human           Human/Agent        Agent executes  Agent writes
    creates          formulates         and reports     synthesis

All agents have access to all MCP tools. There are no role gates -- permissions (pre_research, research, experiment, report, admin) describe what the agent is expected to do, not what tools it can call.

Getting Started

Step 1: Setup MCP

Configure your MCP connection. See references/01-setup.md.

Step 2: Check In

synapse_checkin()

Returns your agent identity, current assignments, and pending work.

Step 3: Follow the Workflow

Task	Guide
Research questions and literature	references/02-research-workflow.md
Running experiments	references/03-experiment-workflow.md
Autonomous experiment proposal	references/04-autonomous-loop.md

Execution Rules

Always check in first -- call synapse_checkin() at session start
Report progress -- use synapse_report_experiment_progress and synapse_add_comment to keep the team informed
Follow the lifecycle -- research questions lead to experiments; experiments produce results and documents
Document decisions -- add comments explaining reasoning on experiments and research questions
Use compute correctly -- get SSH access via synapse_get_node_access_bundle, never assume local key paths

Status Lifecycles

Experiment Status Flow

draft --> pending_review --> pending_start --> in_progress --> completed

Research Question Status Flow

open --> elaborating --> proposal_created --> completed
  \                                            /
   \--> closed <------------------------------/