Back to Details

pre-mortem

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Pre-Mortem Skill

Pre-Mortem 技能

Purpose: Is this plan/spec good enough to implement?
Run 
/council validate
on a plan or spec to get multi-model judgment before committing to implementation.
用途： 该计划/规格说明是否足够成熟可以实施？
运行 
/council validate
命令，在投入实施前对计划或规格说明进行多模型评审。

Quick Start

快速开始

bash
/pre-mortem                                         # validates most recent plan
/pre-mortem path/to/PLAN.md                         # validates specific plan
/pre-mortem --quick path/to/PLAN.md                 # fast inline check, no spawning
/pre-mortem --deep path/to/SPEC.md                  # 4 judges with plan-review preset
/pre-mortem --mixed path/to/PLAN.md                 # cross-vendor (Claude + Codex)
/pre-mortem --preset=architecture path/to/PLAN.md   # architecture-focused review
/pre-mortem --explorers=3 path/to/SPEC.md           # deep investigation of plan
/pre-mortem --debate path/to/PLAN.md                # two-round adversarial review
bash
/pre-mortem                                         # 验证最新的计划
/pre-mortem path/to/PLAN.md                         # 验证指定的计划
/pre-mortem --quick path/to/PLAN.md                 # 快速内联检查，不启动额外进程
/pre-mortem --deep path/to/SPEC.md                  # 4位评审员，使用plan-review预设
/pre-mortem --mixed path/to/PLAN.md                 # 跨厂商模型（Claude + Codex）
/pre-mortem --preset=architecture path/to/PLAN.md   # 聚焦架构的评审
/pre-mortem --explorers=3 path/to/SPEC.md           # 对计划进行深度调研
/pre-mortem --debate path/to/PLAN.md                # 两轮对抗式评审

Execution Steps

执行步骤

Step 1: Find the Plan/Spec

步骤1：查找计划/规格说明

If path provided: Use it directly.
If no path: Find most recent plan:
bash
ls -lt .agents/plans/ 2>/dev/null | head -3
ls -lt .agents/specs/ 2>/dev/null | head -3
Use the most recent file. If nothing found, ask user.
如果提供了路径： 直接使用该路径。
如果未提供路径： 查找最新的计划：
bash
ls -lt .agents/plans/ 2>/dev/null | head -3
ls -lt .agents/specs/ 2>/dev/null | head -3
使用最新的文件。如果未找到任何文件，询问用户。

Step 1a: Search Knowledge Flywheel

步骤1a：搜索知识飞轮

bash
if command -v ao &>/dev/null; then
    ao search "plan validation lessons <goal>" 2>/dev/null | head -10
fi
If ao returns prior plan review findings, include them as context for the council packet. Skip silently if ao is unavailable or returns no results.
bash
if command -v ao &>/dev/null; then
    ao search "plan validation lessons <goal>" 2>/dev/null | head -10
fi
如果ao工具返回了之前的计划评审结果，将其作为上下文包含在评审数据包中。如果ao工具不可用或未返回结果，则跳过此步骤，不提示。

Step 2: Run Council Validation

步骤2：运行评审委员会验证

Run 
/council
with the plan-review preset and always 4 judges (--deep):
/council --deep --preset=plan-review validate <plan-path>
Default (4 judges with plan-review perspectives):
  • missing-requirements
    : What's not in the spec that should be? What questions haven't been asked?
  • feasibility
    : What's technically hard or impossible here? What will take 3x longer than estimated?
  • scope
    : What's unnecessary? What's missing? Where will scope creep?
  • spec-completeness
    : Are boundaries defined? Do conformance checks cover all acceptance criteria? Is the plan mechanically verifiable?
Pre-mortem always uses 4 judges (
--deep
) because plans deserve thorough review. The spec-completeness judge validates SDD patterns; for plans without boundaries/conformance sections, it issues WARN (not FAIL) for backward compatibility.
With --quick (inline, no spawning):
/council --quick validate <plan-path>
Single-agent structured review. Fast sanity check before committing to full council.
With --mixed (cross-vendor):
/council --mixed --preset=plan-review validate <plan-path>
3 Claude + 3 Codex agents for cross-vendor plan validation with plan-review perspectives.
With explicit preset override:
/pre-mortem --preset=architecture path/to/PLAN.md
Explicit 
--preset
overrides the automatic plan-review preset. Uses architecture-focused personas instead.
With explorers:
/council --deep --preset=plan-review --explorers=3 validate <plan-path>
Each judge spawns 3 explorers to investigate aspects of the plan's feasibility against the codebase. Useful for complex migration or refactoring plans.
With debate mode:
/pre-mortem --debate
Enables adversarial two-round review for plan validation. Use for high-stakes plans where multiple valid approaches exist. See 
/council
docs for full --debate details.
使用plan-review预设和固定4位评审员（--deep参数）运行 
/council
命令：
/council --deep --preset=plan-review validate <plan-path>
默认配置（4位评审员，具备plan-review视角）：
  • missing-requirements
    : 规格说明中缺少哪些必要内容？还有哪些问题尚未被提出？
  • feasibility
    : 哪些内容在技术上存在难度或无法实现？哪些部分的耗时会比预估多3倍？
  • scope
    : 哪些内容是不必要的？哪些内容缺失？哪些地方会出现范围蔓延？
  • spec-completeness
    : 边界是否已明确定义？一致性检查是否覆盖了所有验收标准？计划是否可机械验证？
Pre-mortem 始终使用4位评审员（--deep参数），因为计划需要全面评审。spec-completeness评审员会验证SDD模式；对于没有边界/一致性部分的计划，会发出WARN（而非FAIL）以保持向后兼容性。
使用--quick参数（内联检查，不启动额外进程）：
/council --quick validate <plan-path>
单Agent结构化评审。在启动完整评审委员会前进行快速的合理性检查。
使用--mixed参数（跨厂商模型）：
/council --mixed --preset=plan-review validate <plan-path>
3个Claude + 3个Codex Agent，以plan-review视角进行跨厂商计划验证。
使用显式预设覆盖：
/pre-mortem --preset=architecture path/to/PLAN.md
显式的
--preset
参数会自动覆盖plan-review预设，改用聚焦架构的角色进行评审。
使用explorers参数：
/council --deep --preset=plan-review --explorers=3 validate <plan-path>
每位评审员会启动3个探索者Agent，针对计划在代码库中的可行性进行深度调研。适用于复杂的迁移或重构计划。
使用debate模式：
/pre-mortem --debate
为计划验证启用两轮对抗式评审。适用于存在多种可行方案的高风险计划。有关--debate的详细信息，请查看
/council
文档。

Step 3: Interpret Council Verdict

步骤3：解读评审委员会 verdict

Council VerdictPre-Mortem ResultAction
PASSReady to implementProceed
WARNReview concernsAddress warnings or accept risk
FAILNot readyFix issues before implementing
评审委员会 verdictPre-Mortem 结果操作建议
PASS可实施继续推进
WARN需评审关注点解决警告问题或接受风险
FAIL未就绪在实施前修复问题

Step 4: Write Pre-Mortem Report

步骤4：撰写Pre-Mortem报告

Write to: 
.agents/council/YYYY-MM-DD-pre-mortem-<topic>.md
markdown
undefined
保存路径： 
.agents/council/YYYY-MM-DD-pre-mortem-<topic>.md
markdown
undefined

Pre-Mortem: <Topic>

Pre-Mortem: <主题>

Date: YYYY-MM-DD Plan/Spec: <path>
日期： YYYY-MM-DD 计划/规格说明： <路径>

Council Verdict: PASS / WARN / FAIL

评审委员会 verdict: PASS / WARN / FAIL

JudgeVerdictKey Finding
Missing-Requirements......
Feasibility......
Scope......
评审员Verdict关键发现
Missing-Requirements......
Feasibility......
Scope......

Shared Findings

共同发现

  • ...
  • ...

Concerns Raised

提出的关注点

  • ...
  • ...

Recommendation

建议

<council recommendation>
<council recommendation>

Decision Gate

决策节点

[ ] PROCEED - Council passed, ready to implement [ ] ADDRESS - Fix concerns before implementing [ ] RETHINK - Fundamental issues, needs redesign
undefined
[ ] 推进 - 评审通过，可实施 [ ] 整改 - 修复问题后再实施 [ ] 重新设计 - 存在根本性问题，需要重新设计
undefined

Step 5: Record Ratchet Progress

步骤5：记录Ratchet进度

bash
ao ratchet record pre-mortem 2>/dev/null || true
bash
ao ratchet record pre-mortem 2>/dev/null || true

Step 6: Report to User

步骤6：向用户汇报

Tell the user:
  1. Council verdict (PASS/WARN/FAIL)
  2. Key concerns (if any)
  3. Recommendation
  4. Location of pre-mortem report
告知用户以下内容：
  1. 评审委员会的verdict（PASS/WARN/FAIL）
  2. 主要关注点（如有）
  3. 建议
  4. Pre-Mortem报告的位置

Integration with Workflow

工作流集成

/plan epic-123
    │
    ▼
/pre-mortem                    ← You are here
    │
    ├── PASS → /implement
    ├── WARN → Review, then /implement or fix
    └── FAIL → Fix plan, re-run /pre-mortem
/plan epic-123
    │
    ▼
/pre-mortem                    ← 当前步骤
    │
    ├── PASS → /implement
    ├── WARN → 评审后，选择/implement或修复
    └── FAIL → 修复计划，重新运行/pre-mortem

Examples

示例

Validate a Plan

验证计划

bash
/pre-mortem .agents/plans/2026-02-05-auth-system.md
3 judges (missing-requirements, feasibility, scope) review the auth system plan.
bash
/pre-mortem .agents/plans/2026-02-05-auth-system.md
3位评审员（missing-requirements、feasibility、scope）评审认证系统计划。

Cross-Vendor Plan Validation

跨厂商计划验证

bash
/pre-mortem --mixed .agents/plans/2026-02-05-auth-system.md
3 Claude + 3 Codex agents validate the plan with plan-review perspectives.
bash
/pre-mortem --mixed .agents/plans/2026-02-05-auth-system.md
3个Claude + 3个Codex Agent以plan-review视角验证计划。

Architecture-Focused Review

聚焦架构的评审

bash
/pre-mortem --preset=architecture .agents/specs/api-v2-spec.md
3 judges with architecture perspectives (scalability, maintainability, simplicity) review the spec.
bash
/pre-mortem --preset=architecture .agents/specs/api-v2-spec.md
3位具备架构视角（可扩展性、可维护性、简洁性）的评审员评审规格说明。

Auto-Find Recent Plan

自动查找最新计划

bash
/pre-mortem
Finds the most recent plan in 
.agents/plans/
and validates it.
bash
/pre-mortem
在
.agents/plans/
目录中查找最新计划并进行验证。

See Also

另请参阅

  • skills/council/SKILL.md
    — Multi-model validation council
  • skills/plan/SKILL.md
    — Create implementation plans
  • skills/vibe/SKILL.md
    — Validate code after implementation
  • skills/council/SKILL.md
    — 多模型验证评审委员会
  • skills/plan/SKILL.md
    — 创建实施计划
  • skills/vibe/SKILL.md
    — 实施后验证代码