Why is a handoff document necessary for AI coding?

AI sessions have finite context windows. A handoff document compacts thousands of lines of interaction into a concise summary, allowing a new session to understand the project state without re-reading the entire history.

Where are the handoff files saved?

Handoffs are saved in thoughts/shared/handoffs/{session-name}/ using a timestamped filename to keep your project documentation organized.

What metadata does this skill capture?

It automatically captures the session name, researcher name, git commit hash, branch name, repository details, and Braintrust trace IDs for full traceability.

How do I resume work from a handoff?

Once a handoff is created, you can start a new session and use the /resume_handoff command pointing to the generated markdown file in the thoughts/shared/handoffs/ directory.

Does it track failed attempts?

Yes, the skill includes a specific Post-Mortem section to document what worked and what failed, helping future sessions avoid repeating the same mistakes.

Session Handoff Creator

Name: Session Handoff Creator
Author: wildcard

bywildcard

•

Collaboration Tools

Generates comprehensive, structured handoff documents to preserve context and state when transitioning work between AI sessions.

About

Session Handoff Creator is a specialized skill for Claude Code designed to solve the problem of context loss during long-running development tasks. By automating the creation of detailed markdown documents—complete with YAML frontmatter, Braintrust trace IDs, git metadata, and task statuses—this skill allows developers to seamlessly transfer the 'mental model' of a session to a new instance. It captures not just what was done, but why decisions were made, what failed, and exactly where the next agent should pick up, ensuring continuity in complex engineering workflows.

Key Features

Standardized directory structure for easy session resumption
Automated session metadata and git status extraction
Integrated outcome marking for long-term project analytics
Structured post-mortem section for tracking decisions and failures
Braintrust trace ID integration for deep session linking
23 GitHub stars

Use Cases

Providing structured status updates for human developers or team reviews
Resuming complex feature development across multiple AI session limits
Building a historical 'Artifact Index' of project decisions and outcomes

About

Key Features

Standardized directory structure for easy session resumption
Automated session metadata and git status extraction
Integrated outcome marking for long-term project analytics
Structured post-mortem section for tracking decisions and failures
Braintrust trace ID integration for deep session linking
23 GitHub stars

Use Cases

Providing structured status updates for human developers or team reviews
Resuming complex feature development across multiple AI session limits
Building a historical 'Artifact Index' of project decisions and outcomes