What metadata does Drift Start record?

It records the current git SHA, the branch name, an ISO 8601 timestamp, the ticket ID (e.g., PROJ-123), and the description of the planned work.

Is a Jira account required to use this skill?

No, Jira integration is optional. While it can automatically fetch ticket summaries, you can also provide ticket IDs and descriptions manually.

Where is the drift session data stored?

The data is stored locally in your project's .drift/ directory. The skill automatically adds this directory to your .gitignore to prevent local session data from being committed.

How does the skill handle Jira credentials safely?

It uses a tiered resolution order, checking environment variables, local .env files, drift-specific .env files, or global config files before prompting the user.

What happens if I start a new session while one is already active?

The skill will detect the existing session, show you the details, and ask for confirmation before overwriting it with a new starting point.

Drift Start

Name: Drift Start
Author: Hamad-Center

byHamad-Center

•

开发者工具

Initializes a code drift tracking session by recording the current git state and project metadata to track implementation evolution.

Drift Start is a specialized utility for Claude Code designed to baseline development tasks before implementation begins. By capturing the current git SHA, branch, and associated ticket identifiers, it creates a structured reference point to monitor how code diverges from the initial plan. The skill automates the creation of local configuration files, manages .gitignore settings, and provides seamless Jira integration to fetch ticket summaries and descriptions, ensuring that every coding session is contextualized and documented for future analysis.

主要功能

01Automated .drift directory initialization and .gitignore configuration

023 GitHub stars

03Automatic git repository verification and SHA recording

04Intelligent Jira ticket metadata fetching and synchronization

05Robust credential resolution from environment variables, .env files, or global configs

06Session history management with a rolling 20-entry limit

使用场景

01Linking Claude Code sessions to Jira tickets for improved traceability

02Monitoring implementation drift during collaborative development tasks

03Baselining a codebase before starting a complex feature or refactor

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add hamad-center/drift-skills drift-start

For use in Claude.ai and ChatGPT

Download Skill