elsa-whats-next

name	elsa-whats-next
description	Find and rank the next Elsa-brain work unit from unfinished work, reports, specs, and source markers. Use when a user asks what is next, what is unfinished, what is unratified, what is weakly implemented, or which unit should be planned next in elsa-foundation.
argument-hint	Optional focus area
compatibility	Requires elsa-foundation docs and AGENTS.md
metadata	{"author":"elsa-foundation","source":"docs/skills/catalog.md#whats-next--unfinished-work"}
user-invocable	true
disable-model-invocation	false

User Input

$ARGUMENTS

Outline

1. Read AGENTS.md, especially source-of-truth layers and the program goals / drift guard.
2. Read docs/skills/catalog.md#whats-next--unfinished-work.
3. Inspect docs/program-goals/ to identify the current program-goal state when it affects ranking or is unclear: a named bucket, none/free-flow, or temporarily unknown/not-assessed.
4. Read .agent-prefs/session-execution-model.md if it exists.
5. Read docs/reports/unfinished-work.md as inventory/evidence, not an active queue.
6. Refresh context with the marker search from unfinished-work.md when the user asks for current state or when freshness is uncertain.
7. Rank candidates by current program-goal state when applicable, user intent, severity, and unblock value.
8. Return a concise drift check, top candidates, recommendation, source-of-truth layer, program-goal route, program-goal state when used, and execution route. If the recommended unit requires substantial planning or implementation and the local session preference favors fresh workers or ask-each-time, ask whether to plan/execute here or prepare a reviewed handoff prompt for a fresh agent/thread.

Do not implement from this skill. If the user approves a unit, hand off to elsa-work-unit-planner or the relevant implementation skill.