serena-init

Initializes Serena MCP projects and writes onboarding memory files for new codebases.

Updated Apr 15, 2026
One-click install
npx skills add https://github.com/shoshoavi/agentic_worflows --skill serena-init-shoshoavi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: serena-init
Source: https://github.com/shoshoavi/agentic_worflows/tree/main/cursor/skills/serena-init
Command: npx skills add https://github.com/shoshoavi/agentic_worflows --skill serena-init-shoshoavi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Setting up Serena MCP on a new repository requires manual activation, project detection, and knowledge capture. This Skill walks through the full initialization flow so semantic code navigation works immediately on any new codebase. ## Core Features & Use Cases - Project Activation: Checks the current Serena configuration and activates the workspace by absolute path with automatic language detection. - Guided Onboarding: Reads key project files (README, package manifests, entry points) and writes structured memory files covering project overview, suggested commands, and code style conventions. - Setup Verification: Confirms memories were created and provides troubleshooting for common errors like wrong project paths or invalid tool arguments. - Use Case: You clone a new repository and want Serena's semantic editing tools available. Invoke this Skill to activate the project, generate memory files, and verify the setup in one pass. ## Quick Start Ask the assistant to initialize Serena for the current workspace and run onboarding to create the project memory files.

Frequently Asked Questions about serena-init

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up Serena MCP for a new project?▼

Activate the project by calling activate_project with the full absolute path to the repository root. Serena auto-detects the programming language, then you can run onboarding to generate memory files describing the project.

How to onboard a codebase to Serena?▼

Run the onboarding tool, then read key files like README, package.json, or requirements.txt to understand the project. Write memory files for project_overview, suggested_commands, and code_style_conventions using write_memory.

Why does Serena say project not found?▼

This error occurs when you pass a relative path instead of a full absolute path, or when the directory does not exist or contains no code files. Always use the complete workspace root path when calling activate_project.

What memory files should I create during Serena onboarding?▼

Create three core memories: project_overview covering purpose, tech stack, and structure; suggested_commands for setup, run, test, and build commands; and code_style_conventions for naming and patterns used in the codebase.

Why does write_memory fail with a validation error?▼

The write_memory tool expects the argument memory_file_name, not name. Check the tool's JSON schema under the Serena MCP tools directory to confirm the exact parameter names before calling it.