Add optional /goal candidate output to craft-handoff

[sungjunlee]

Problem

craft-handoff produces a strong continuation prompt, but some handoffs naturally lead into long, verifiable, multi-turn work. In those cases, users often want a /goal they can review and activate in the next session.

The current handoff flow has no explicit rule for producing a safe optional goal candidate.

Proposed change

Update skills/craft-handoff/SKILL.md so it may include an optional /goal candidate when the next task is all of the following:

• durable enough to need multiple turns
• has an observable end state
• has a concrete verification surface such as a command, report, artifact, or transcript-visible check
• has meaningful constraints or non-goals
• has a reasonable budget or blocked-stop condition

The candidate should be clearly labeled as review-before-activating and must not be auto-run.

Suggested shape:

Optional goal candidate, review before activating:

/goal Continue from the handoff prompt until <observable end state>, verified by <command/artifact output visible in transcript>, while preserving <constraints>. Stop after <N> turns or if <blocked condition>.

Acceptance criteria

• Normal handoffs remain unchanged when the next step is small, vague, or not verifiable.
• The optional candidate appears only after the primary handoff prompt instructions, not inside the rich doc as an operative task.
• The wording makes activation an explicit user action.
• The candidate includes evidence, constraints, and a stop condition.
• Claude-specific auto-load guidance remains separate from cross-agent /goal guidance.

Non-goals

• Do not create a new craft-goal skill yet.
• Do not auto-run /goal from craft-handoff.
• Do not write repo-local goals/ files from handoff by default.

[coderabbitai]

🔗 Related PRs

#3 - Add skill-radar curation layer + craft-skill-spec meta-skill [merged]
#4 - refactor(craft-handoff): per-session pending files, worktree-scoped auto-load [merged]
#5 - craft-handoff: pair per-session prompt with per-project rich doc [merged]
#7 - docs(craft-prompt): add /goal completion-condition reference [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!