How do I install the Elsa Work Unit Planner skill?

Run npx skillmds@latest add elsa-workflows/elsa-work-unit-planner in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Elsa Work Unit Planner safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Elsa Work Unit Planner?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Elsa Work Unit Planner free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Elsa Work Unit Planner?

elsa-workflows (@elsa-workflows) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Elsa Work Unit Planner by elsa-workflows

Name: Elsa Work Unit Planner
Author: elsa-workflows

Elsa Work Unit Planner

Turn Elsa-brain findings into an architecture-first or Speckit-ready work-unit plan. Use when a report finding, deferred decision, weak implementation, docs/maps/skills gap, or user-selected next unit needs scope, success criteria, source-of-truth layer, and exact files before implementation.

File contents

User Input

$ARGUMENTS

Outline

Read AGENTS.md for source-of-truth layers and the program goals / drift guard.
Read docs/skills/catalog.md#work-unit-planner and docs/skills/catalog.md#promote-finding-to-work-unit.
Inspect docs/program-goals/ and identify whether the work belongs in a named bucket or should remain none/free-flow.
Read the source report/spec/constitution section behind the finding.
Classify the unit as architecture development, feature development, codebase verification, docs/maps/skills work, or code.
Define goal, success criteria, in scope, out of scope, source-of-truth layer, program-goal route, program-goal state when used, affected gates/maps/docs, tests, and review points.
End with a <proposed_plan> block listing exact files to create or update, including likely follow-through obligations such as tests, extension-point catalogs, generated-map refreshes, or docs updates.

Do not run Speckit or edit code unless the user explicitly approves implementation. After a unit is approved, required follow-through obligations are part of completing that unit unless they introduce a new architecture decision.

Frequently asked questions