Teach craft-prompt to gather /goal contract inputs

[sungjunlee]

Problem

craft-prompt already references /goal completion conditions, but its main workflow does not explicitly gather the fields that make a goal safe to run for many turns.

This can lead to weak goal prompts that name the task but omit completion evidence, non-goals, scope, budget, or blocked-stop behavior.

Proposed change

Update skills/craft-prompt/SKILL.md so Step 2 includes a Goal condition / Goal spec raw-material type.

It should gather, when relevant:

• outcome: what must be true when complete
• evidence: command, artifact, report, source, or transcript-visible proof
• constraints: behavior, APIs, files, compatibility, style, data, or performance that must not regress
• non-goals: what the agent should avoid even if it seems useful
• scope: allowed files, repos, tools, network, credentials, and data sources
• budget: turn, time, token, command, or cost cap
• blocked stop condition: when to stop and what to report

Acceptance criteria

• craft-prompt explicitly recognizes /goal work as a distinct prompt type.
• The workflow remains cross-agent and avoids Codex-only wording in the main spine.
• The change is small and points detailed guidance to references/goal-conditions.md rather than bloating SKILL.md.
• Korean trigger wording such as goal 조건 still makes sense with the updated description.

Notes

This came from reviewing the /goal guidance against goal-setter, Codex docs, Claude Code docs, craft-survey, craft-critique, plan-eng-review, and simplify perspectives.

[coderabbitai]

🔗 Related PRs

#3 - Add skill-radar curation layer + craft-skill-spec meta-skill [merged]
#5 - craft-handoff: pair per-session prompt with per-project rich doc [merged]
#7 - docs(craft-prompt): add /goal completion-condition reference [merged]
#22 - [codex] implement CraftKit review backlog [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!