Work Item
Productivitywork-itemWork Item is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Update the current work item displayed in the status line. Use when you want to set or analyze what task is being worked on. Triggers on "/work-item", "update work item", "set current task", or "what am I working on".
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/data/work-item-tracker/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/work-item, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Work Item Tracker
Updates the current work item displayed in the Claude Code status line.
Usage
/work-item- Analyze conversation and auto-detect current task/work-item "Custom description"- Set a specific work item manually
How It Works
The work item is displayed in the status line between the git branch and model name, giving you constant visibility into what task Claude is tracking.
Instructions
When this skill is invoked:
If an argument is provided (e.g., /work-item "Adding dark mode")
- Extract the argument text (remove surrounding quotes if present)
- Write it directly to the cache file:
echo "Adding dark mode" > ~/.claude/work-item-cache.txt - Confirm to the user: "Work item updated to: Adding dark mode"
If no argument is provided (e.g., just /work-item)
- Analyze the recent conversation context to determine the current task
- Generate a concise summary (max 25 characters) that describes what's being worked on
- Use action-oriented phrasing like:
- "Adding user auth"
- "Fixing API bug"
- "Refactoring tests"
- "Implementing search"
- Write the summary to the cache file:
echo "Your summary here" > ~/.claude/work-item-cache.txt - Confirm to the user: "Work item updated to: [summary]"
Cache File Location
~/.claude/work-item-cache.txt
This file is:
- Initialized to "New session" when a Claude Code session starts
- Updated when this skill is invoked
- Read by the statusLine command to display the current work item
Example Status Line Output
orient | dev | Adding user auth | opus | ctx: 15%
Tips
- Run
/work-itemperiodically when switching tasks - Use custom descriptions when auto-detection doesn't capture the right context
- Keep descriptions short - they're truncated to 30 characters in the status line