hunt

Compose a timed coding shift from a ready catalog of work entries for review or direct execution.

63|3|Updated Jul 14, 2026
One-click install
npx skills add https://github.com/orwa-mahmoud/nightshift --skill hunt-orwa-mahmoud
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hunt
Source: https://github.com/orwa-mahmoud/nightshift/tree/main/plugins/nightshift/skills/hunt
Command: npx skills add https://github.com/orwa-mahmoud/nightshift --skill hunt-orwa-mahmoud

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long autonomous coding sessions often stall because the agent lacks a concrete, time-boxed work plan. This Skill turns a loose objective or a ready-made catalog of shift entries into a structured, deadline-bound shift that an agent can execute and a human can review. ## Core Features & Use Cases - Catalog-driven composition: Discovers shift entries (defect hunts, coverage hunts, lint debt, documentation drift) from the installed plugin catalog and combines compatible entries under one deadline. - Guided or automatic modes: Lets the owner pick entries from a menu, or composes the shift automatically from a complete natural-language objective with hours and intent. - Review-first or run-direct execution: Prints a no-write preview of the shift for approval, or cuts and arms the shift immediately with a deadline, gate, and watchman. - Use Case: Tell the agent "use the next 8 hours clearing lint and test debt" and it composes the matching catalog entries, sets the tooling policy, writes the work order, and starts the shift without further prompts. ## Quick Start Ask the agent to compose and run an eight-hour nightshift shift that clears lint and test debt in the current project.

Frequently Asked Questions about hunt

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

FAQPage Schema
How do I start a long autonomous coding session with a deadline?

Give the agent a complete prompt with an objective, hours, and run intent, such as "8 hours clear lint and test debt". The skill composes matching catalog entries, sets the tooling policy, writes the work order, and starts the shift with a deadline and gate.

What is the difference between guided and automatic shift composition?

Guided mode lists each catalog entry with its ending so the owner picks the work, while automatic mode inspects the work target and composes entries that support the stated objective. Both modes can run review-first or run-direct.

Can I review the shift plan before the agent starts working?

Yes, choose review-first mode and the skill prints the exact items, evidence, hours, and ending as a no-write preview. Nothing is written or armed until you approve, and the clock starts only at the cut.

Does the hunt skill work with Claude Code, Codex, and Cursor?

Yes, it resolves the plugin root per host and runs every command through the shared ns runtime, which handles host and workspace resolution. Native Windows uses the PowerShell entry point with the same verbs.

Which shift entries cannot run in artifact mode?

Repository-shaped entries are refused in artifact mode: defect hunt, documentation drift, TODO and FIXME debt, coverage hunt, tooling quality-debt entries, and the GitHub issue-hunt entry. Only existing-tools is a valid tooling policy there.