菜单
Map what we're optimizing and what constraints we treat as real. Use before jumping to solutions, when hitting repeated blockers, or when patches keep accumulating.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Clarify the outcome you want - a change in user behavior, not a feature shipped. Use at the start of any work to ground the session in strategic intent.
Devil's advocate. Seek contrary evidence before locking in. Use when about to make a significant decision, when confidence is high but stakes are higher, or when people are converging too quickly.
Human-led curation of accumulated metis and guardrails. Surface patterns across sessions, propose what to promote, compact, or dismiss. Use after multiple sessions, before a new phase, or when search results feel noisy.
Do the work. Pre-flight, build, detect drift, salvage if needed. Use when you have a clear aim and are ready to implement.
Define the framing of a problem. Change the statement, change the solution space. Use when starting work, when solutions feel wrong, or when you suspect an X-Y problem.
Check work and detect drift before committing. A second opinion that catches misalignment early. Use at natural pause points, before PRs, or when something feels off.
| name | problem-space |
| description | Map what we're optimizing and what constraints we treat as real. Use before jumping to solutions, when hitting repeated blockers, or when patches keep accumulating. |
Map the terrain where solutions live: objective, constraints, systems, assumptions, and what would prove the map wrong.
Problem space precedes solution space. Jump to code too early and you build the wrong thing fast.
Skip when: The problem is well-understood and you are in execution.
"We are optimizing for [outcome]."
Name the behavior change, metric, or problem disappearance that indicates success. Features are not objectives.
List what is treated as fixed:
Constraint: [boundary]
Type: [hard | soft | assumed]
Reason: [why it exists]
Questioning: [could this be false?]
Hard constraints do not bend. Soft and assumed constraints can be questioned. Do not talk yourself out of real constraints because implementation feels easy.
Capture only terrain that changes decisions:
If RNA/OH context tools are available, surface guardrails and metis as candidates with provenance; the human selects what carries forward.
Connect objective, terrain, constraints, assumptions, evidence quality, and open questions. Facts with no decision impact are noise.
Make hidden assumptions visible. Check for X-Y mismatch. Name frame-stress signals: evidence that constraints, terrain, assumptions, or precedents were wrong.
If constraints, assumptions, or open questions remain implicit, you are not ready for /solution-space.
## Problem Space
**Scope:** [what area this covers]
### Objective
[Outcome, not feature]
### Constraints
| Constraint | Type | Reason | Question? |
|------------|------|--------|-----------|
| [boundary] | hard/soft/assumed | [why] | [could this be false?] |
### Terrain
- **Systems:** [what's involved]
- **Stakeholders:** [who's affected]
- **Blast radius:** [what breaks if wrong]
- **Precedents/metis:** [existing local knowledge]
### Situation Model
- **Explains:** [why these elements matter]
- **May hide:** [blind spots]
- **Evidence quality:** [observed / inferred / assumed]
### Assumptions and Open Questions
- [assumption or unknown] — [risk if false / when to answer]
### Frame-Stress Signals
- [evidence that should trigger problem-space revision]
### Ready for Solution Space?
[yes/no] - [why or what's missing]
If a session file is in use, read prior Aim, then replace or append ## Problem Space. Later phases carry forward explicit constraints, assumptions, evidence quality, frame-stress signals, and open questions.