skills.rest
shenxingy
shenxingyCOMMUNITY
Explore all skills by this author →
Website𝕏@alexshen_devGitHub

next

Recommends next project task from git history and TODO docs with optional interview.

8|5|Updated Feb 18, 2026
Search Tags:#project planning#prioritization#next steps#agent workflows#roadmap alignment#git context#execution vs discovery
One-click install
npx skills add https://github.com/shenxingy/Clade --skill next-shenxingy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: next
Source: https://github.com/shenxingy/Clade/tree/main/configs/skills/next
Command: npx skills add https://github.com/shenxingy/Clade --skill next-shenxingy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you decide what to work on next by turning your existing project docs and git state into a prioritized “next move,” with an optional deeper interview when priorities are unclear.

Core Features & Use Cases

  • Fast mode one-shot prioritization: Recommends a top pick (and optional runner-up) by reading your project’s TODO/progress/goals plus recent git activity, then outputs a short plan for quick execution.
  • Deep mode multi-round discovery: Runs a structured interview (framing check + multiple rounds) to surface real pain, momentum, horizon alignment, and whether tasks are discovery vs execution before synthesizing a robust plan.
  • Effort-aware estimates: Translates human effort language into Claude Code/agent time equivalents so the recommendation matches how long autonomous runs typically take.
  • Safeguards against bad execution: Explicitly distinguishes discovery tasks (needs a spike) from execution tasks (can be run directly) to reduce stalled autonomous work.

Quick Start

Ask your AI to run next in fast mode by saying: /next.

Frequently Asked Questions about next

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

FAQPage Schema
How do I decide what project task to work on next?▼

To decide what project task to work on next, synthesize existing project docs and recent git activity to generate a prioritized top pick. This approach reads your TODO files and momentum to recommend a quick execution plan.

What is the best way to prioritize tasks when project goals are unclear?▼

Prioritizing tasks with unclear goals requires a deep mode multi-round discovery interview to surface real pain and horizon alignment. This structured framing distinguishes discovery spikes from direct execution work.

How do I prioritize backlog items using current git context?▼

Prioritize backlog items using current git context by reading recent commits alongside project progress documents. Fast mode one-shot prioritization outputs a top pick and runner-up based on this synthesized momentum.

Can I estimate agent time for autonomous project workflows?▼

You can estimate agent time for autonomous project workflows by translating human effort language into Claude Code time equivalents. Effort-aware estimates match recommendations to how long autonomous runs typically take.

When should I distinguish discovery tasks from execution tasks?▼

Distinguish discovery tasks from execution tasks to prevent stalled autonomous work and bad execution. Explicitly framing tasks as discovery spikes versus direct execution safeguards against autonomous agents running stalled work.

Related Skills

Explore More Product & Management Skills→
dokdo2013dokdo2013
1

todo-review

Verify todo implementation status by inspecting repository files and git history.

Community
Intermediate
akshatdaltonakshatdalton

local-schedule

Schedules Claude Code tasks via crontab and claude -p with ntfy.sh notifications.

Community
Intermediate
MaxKlat29MaxKlat29

product-strategy-session

Guide product teams through a six-phase strategy process with workshops and validation.

Community
Advanced
dralgorhythmdralgorhythm
111

estimating-work

Estimate development effort using story points and T-shirt sizing.

Community
Intermediate