objective-next

Recommends and serializes the next semantic step for an active Objective as a decision packet.

2|Updated Apr 2, 2026
One-click install
npx skills add https://github.com/nseng-ai/ns --skill objective-next-nseng-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: objective-next
Source: https://github.com/nseng-ai/ns/tree/main/skills/incubating/objectives/objective-next
Command: npx skills add https://github.com/nseng-ai/ns --skill objective-next-nseng-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? When working on long-running, multi-session Objectives tracked in a repository, it is hard to decide what to do next, whether the tracking record is stale, and whether work can be executed directly. This Skill resolves one active Objective, checks its freshness against git evidence, and produces a decision packet ending in a ready-to-run proposed prompt or an explicit decline with routing. ## Core Features & Use Cases - Staleness-aware recommendation: Runs a read-only staleness check via ns objective exec staleness-check and routes through objective-update before recommending work when tracking lags real progress. - Decision packet output: Emits a scannable packet — basis, work-left estimate, alternatives, and a self-contained proposed prompt that can run cold in a fresh session or subagent. - Blocked-Objective routing: Traverses edge annotations to find the counterpart Objective that would unblock the selected one and shapes the recommendation accordingly. - Confirmed execution: Supports durable ## Runner Policy and recommendation-continuation bases for executing bounded slices after explicit confirmation. - Use Case: You return to a repo after a week, ask what to do next on your current Objective, and receive a one-line basis, a work-left estimate, and a copy-pasteable prompt scoped to the exact next roadmap slice. ## Quick Start Ask the agent to run objective-next for your active Objective slug to get the next recommended step as a ready-to-run prompt.

Frequently Asked Questions about objective-next

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

FAQPage Schema
How do I decide what to work on next in a long-running project objective?

Run objective-next on the active Objective slug. It reads the objective record, roadmap, and updates, checks staleness against git evidence, then emits a decision packet whose final element is a self-contained proposed prompt for the next semantic step.

How does objective-next handle stale objective tracking?

It runs ns objective exec staleness-check to compare branch-diff and uncommitted changes against the Objective record. If meaningful progress is unrecorded, it routes through objective-update first, then rechecks before recommending work.

Can objective-next execute work directly instead of just recommending it?

Yes, when a durable Runner Policy or Definition of Progress permits it, or when the user confirms a current-session recommendation. Otherwise it emits a decision packet or execution preview and waits for explicit confirmation.

What happens when an Objective is blocked?

The skill reads the blocked sentence and the record's edges to find the counterpart Objective that would unblock it. It then recommends advancing that counterpart or non-gated work, and routes stale blockers through objective-update.

When does objective-next decline to propose work?

It emits a Declined packet when only validation-only rows remain, when no active semantic work remains, or when an external gate must clear. The decline names the reason and routes to objective-update, a counterpart Objective, or objective-close.