ln-813-optimization-plan-validator

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Paths: File paths (
shared/
,
references/
,
../ln-*
) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root.

路径： 文件路径（
shared/
、
references/
、
../ln-*
）是相对于技能仓库根目录的。如果在当前工作目录（CWD）未找到，找到此SKILL.md所在目录并向上一级即为仓库根目录。

ln-813-optimization-plan-validator

Type: L3 Worker Category: 8XX Optimization

Validates optimization plan (performance_map + hypotheses + context) via parallel agent review before committing to code changes. Catches feasibility issues, missing hypotheses, and incorrect conflict mappings before the strike.

类型： L3 Worker 分类： 8XX 优化

在提交代码变更前，通过多Agent并行审核验证优化方案（performance_map + 假设 + 上下文），在执行前发现可行性问题、缺失的假设以及错误的冲突映射。

Overview

概述

Aspect	Details
Input	`.optimization/{slug}/context.md` (performance_map, hypotheses, suspicion_stack)
Output	Verdict (GO / GO_WITH_CONCERNS / NO_GO), corrected context.md, agent feedback summary
Pattern	Parallel agent review (Codex + Gemini) + own feasibility check → merge → verdict

方面	详情
输入	`.optimization/{slug}/context.md` （performance_map、假设、疑点栈）
输出	结论（GO / GO_WITH_CONCERNS / NO_GO）、修正后的context.md、Agent反馈摘要
模式	多Agent并行审核（Codex + Gemini）+ 自主可行性检查 → 合并结果 → 给出结论

Workflow

工作流

Phases: Load Context + Health Check → Materialize for Agents → Launch Agents → Feasibility Check → Merge + Debate → Verdict

阶段： 加载上下文 + 健康检查 → 为Agent准备上下文 → 启动Agent → 可行性检查 → 合并反馈 + 研讨 → 给出结论

Phase 0: Load Context + Health Check

阶段0：加载上下文 + 健康检查

MANDATORY READ: Load

shared/references/agent_review_workflow.md

MANDATORY READ: Load

shared/references/agent_delegation_pattern.md

必读内容： 加载

shared/references/agent_review_workflow.md

必读内容： 加载

shared/references/agent_delegation_pattern.md

Slug Resolution

Slug解析

If invoked via Agent with contextStore containing
```
slug
```
— use directly.
If invoked standalone — ask user for
```
.optimization/
```
target directory or scan for single slug.

如果由包含
```
slug
```
的contextStore的Agent调用 — 直接使用该slug。
如果独立调用 — 询问用户
```
.optimization/
```
目标目录或扫描单个slug。

Step 1: Load Context

步骤1：加载上下文

Read

.optimization/{slug}/context.md

from project root. Verify required sections present:

Section	Required	Verify
Performance Map	Yes	`performance_map.baseline` has measurements
Hypotheses	Yes	At least 1 hypothesis with `files_to_modify`
Suspicion Stack	Yes	At least 1 confirmed suspicion
Test Command	Yes	Non-empty `test_command`

If missing → Block: "context.md incomplete — run profiler and researcher first."

从项目根目录读取

.optimization/{slug}/context.md

，验证是否存在所需章节：

章节	是否必填	验证内容
性能地图	是	`performance_map.baseline` 包含测量数据
假设	是	至少1个带有 `files_to_modify` 的假设
疑点栈	是	至少1个已确认的疑点
测试命令	是	`test_command` 非空

若缺失 → 拦截："context.md不完整 — 请先运行分析器和调研工具。"

Step 2: Agent Health Check

步骤2：Agent健康检查

python shared/agents/agent_runner.py --health-check

0 agents available →
```
agents_launched = SKIPPED
```
, proceed with own feasibility check only
Agents available → continue to Phase 1

python shared/agents/agent_runner.py --health-check

无可用Agent →
```
agents_launched = SKIPPED
```
，仅执行自主可行性检查
有可用Agent → 继续阶段1

Phase 1: Materialize Context for Agents

阶段1：为Agent准备上下文

Prepare context for external agents (they cannot read

.optimization/

directly):

Ensure
```
.agent-review/
```
directory exists (with
```
.gitignore
```
containing
```
*
```
)

Copy

.optimization/{slug}/context.md

→

.agent-review/context/{id}_optimization_plan.md

Build per-agent prompts per
```
agent_review_workflow.md
```
Step: Build Prompt (steps 1-9). Use
```
review_base.md
```
+
```
modes/plan_review.md
```

为外部Agent准备上下文（它们无法直接读取

.optimization/

）：

确保
```
.agent-review/
```
目录存在（包含
```
*
```
的
```
.gitignore
```
文件）

复制

.optimization/{slug}/context.md

→

.agent-review/context/{id}_optimization_plan.md

根据
```
agent_review_workflow.md
```
的“构建提示”步骤（步骤1-9）为每个Agent构建提示。使用
```
review_base.md
```
+
```
modes/plan_review.md
```

Optimization-Specific Focus Areas

优化专项关注点

Replace default

{focus_areas}

in prompt with:

MANDATORY READ: Load optimization_review_focus.md

Save per-agent prompts to

.agent-review/{agent}/{id}_optimization_review_prompt.md

将提示中的默认

{focus_areas}

替换为：

必读内容： 加载 optimization_review_focus.md

将每个Agent的提示保存至

.agent-review/{agent}/{id}_optimization_review_prompt.md

Phase 2: Launch Agents (Background)

阶段2：启动Agent（后台运行）

Launch BOTH agents as background Bash tasks:

bash

python shared/agents/agent_runner.py \
  --agent codex-review \
  --prompt-file .agent-review/codex/{id}_optimization_review_prompt.md \
  --output-file .agent-review/codex/{id}_optimization_review.md \
  --cwd {project_root}

python shared/agents/agent_runner.py \
  --agent gemini-review \
  --prompt-file .agent-review/gemini/{id}_optimization_review_prompt.md \
  --output-file .agent-review/gemini/{id}_optimization_review.md \
  --cwd {project_root}

Both run in background (

run_in_background=true

). Proceed to Phase 3 while agents work.

以后台Bash任务启动两个Agent：

bash

python shared/agents/agent_runner.py \
  --agent codex-review \
  --prompt-file .agent-review/codex/{id}_optimization_review_prompt.md \
  --output-file .agent-review/codex/{id}_optimization_review.md \
  --cwd {project_root}

python shared/agents/agent_runner.py \
  --agent gemini-review \
  --prompt-file .agent-review/gemini/{id}_optimization_review_prompt.md \
  --output-file .agent-review/gemini/{id}_optimization_review.md \
  --cwd {project_root}

两者均后台运行（

run_in_background=true

），在Agent运行期间继续阶段3。

Phase 3: Own Feasibility Check (while agents run)

阶段3：自主可行性检查（Agent运行期间）

Perform independent validation of the optimization plan:

Check	How	Fail Action
Files exist	For each hypothesis: verify every file in `files_to_modify` exists	Flag hypothesis as INVALID
No file conflicts	Check uncontested hypotheses don't modify same file lines	Flag overlap as CONCERN
Suspicion coverage	Cross-reference `suspicion_stack` (confirmed) with hypotheses	Flag uncovered suspicions as MISSING_HYPOTHESIS
Evidence backing	Each hypothesis should trace to a profiler finding or research source	Flag unsupported as WEAK_EVIDENCE
Conflicts correct	Verify `conflicts_with` mappings make sense (H1 really makes H3 unnecessary?)	Flag incorrect as BAD_CONFLICT
Fix Hierarchy	Verify hypotheses ordered Configuration→...→Removal. Flag if top hypothesis is level 4-5	CONCERN: "config-level fix may be available"
Removal guard	Any "remove feature" hypothesis MUST have paired "optimize feature" alternative	CONCERN: "removal without optimization alternative"
Assumption verification	Each hypothesis's premises — verified by profiler data or just assumed?	Flag: "assumption not verified: {premise}"
Depth check	Did profiler go inside all accessible slow services? Check performance_map for surface-level entries	CONCERN: "service X profiled at surface level only"

对优化方案执行独立验证：

检查项	检查方式	失败处理
文件存在性	针对每个假设：验证 `files_to_modify` 中的所有文件是否存在	将假设标记为INVALID
无文件冲突	检查无争议的假设是否修改同一文件的相同行	将重叠标记为CONCERN
疑点覆盖	交叉对比已确认的 `suspicion_stack` 与假设	将未覆盖的疑点标记为MISSING_HYPOTHESIS
证据支持	每个假设应能追溯到分析器结果或调研来源	将无支持的假设标记为WEAK_EVIDENCE
冲突映射正确性	验证 `conflicts_with` 映射是否合理（H1是否真的使H3变得不必要？）	将错误映射标记为BAD_CONFLICT
修复层级	验证假设是否按“配置→...→移除”排序。若顶级假设为4-5级则标记	关注点："可能存在配置级修复方案"
移除防护	任何“移除功能”的假设必须配有“优化功能”的替代方案	关注点："仅移除未提供优化替代方案"
假设验证	每个假设的前提 — 是由分析器数据验证还是仅为假设？	标记："假设未验证：{premise}"
深度检查	分析器是否深入了所有可访问的慢服务？检查performance_map是否仅为表层条目	关注点："仅对服务X进行了表层分析"

Output

输出

feasibility_result:
  valid_hypotheses: [H1, H2, H4]
  invalid_hypotheses: [{id: H3, reason: "file not found: src/cache.py"}]
  concerns: [{type: "file_overlap", detail: "H1 and H2 both modify src/api.py"}]
  missing: [{suspicion: "N+1 in loop at handler.py:45", note: "no hypothesis addresses this"}]

feasibility_result:
  valid_hypotheses: [H1, H2, H4]
  invalid_hypotheses: [{id: H3, reason: "file not found: src/cache.py"}]
  concerns: [{type: "file_overlap", detail: "H1 and H2 both modify src/api.py"}]
  missing: [{suspicion: "N+1 in loop at handler.py:45", note: "no hypothesis addresses this"}]

Phase 4: Merge Agent Feedback + Debate

阶段4：合并Agent反馈 + 研讨

Wait for agent results, then merge per

shared/references/agent_review_workflow.md

Parse agent suggestions from both result files
Merge with own feasibility findings (Phase 3)
For EACH suggestion: dedup → evaluate → AGREE or DISAGREE (debate per shared workflow)
Apply accepted corrections directly to
```
.optimization/{slug}/context.md
```
:
- Remove invalid hypotheses
- Add warnings to concerns
- Adjust
```
conflicts_with
```
  if agents found errors
- Add missing hypotheses if agents identified gaps

Save review summary →

.agent-review/review_history.md

Display:

"Agent Review: codex ({accepted}/{total}), gemini ({accepted}/{total}), {N} corrections applied"

等待Agent结果，然后根据

shared/references/agent_review_workflow.md

合并：

解析两个结果文件中的Agent建议
与阶段3的自主可行性检查结果合并
对每个建议：去重 → 评估 → 同意或不同意（按照共享工作流研讨）
将已接受的修正直接应用到
```
.optimization/{slug}/context.md
```
：
- 移除无效假设
- 为关注点添加警告
- 若Agent发现错误则调整
```
conflicts_with
```
- 若Agent发现缺口则添加缺失的假设

将审核摘要保存至 →

.agent-review/review_history.md

显示：

"Agent审核：codex（{已接受}/{总数}），gemini（{已接受}/{总数}），已应用{N}项修正"

Phase 5: Verdict

阶段5：结论

Verdict	Condition
GO	All hypotheses valid, no critical issues, agents agree plan is feasible
GO_WITH_CONCERNS	Minor issues found and documented as warnings in context.md. Safe to proceed
NO_GO	Critical feasibility issue (files missing, fundamental approach flaw, both agents reject)

结论	条件
GO（执行）	所有假设有效，无关键问题，Agents一致认为方案可行
GO_WITH_CONCERNS（带关注点执行）	发现轻微问题并在context.md中记录为警告，可安全执行
NO_GO（不执行）	存在严重可行性问题（文件缺失、核心方法缺陷、两个Agent均拒绝）

Output

输出

validation_result:
  verdict: "GO" | "GO_WITH_CONCERNS" | "NO_GO"
  corrections_applied: <number>
  hypotheses_removed: [<ids>]
  hypotheses_added: [<ids>]
  concerns: [<list>]
  agent_summary: "codex: PLAN_ACCEPTABLE, gemini: SUGGESTIONS (2 accepted)"

Return verdict to coordinator. On NO_GO: coordinator presents issues to user.

validation_result:
  verdict: "GO" | "GO_WITH_CONCERNS" | "NO_GO"
  corrections_applied: <number>
  hypotheses_removed: [<ids>]
  hypotheses_added: [<ids>]
  concerns: [<list>]
  agent_summary: "codex: PLAN_ACCEPTABLE, gemini: SUGGESTIONS (2 accepted)"

将结论返回给协调器。若为NO_GO：协调器向用户展示问题。

Error Handling

错误处理

Error	Recovery
Context file missing	Block: "run profiler and researcher first"
Both agents unavailable	Proceed with own feasibility check only (reduced confidence)
Agent timeout	Use results from available agent + own check
Context file malformed	Block: "context.md missing required sections"

错误	恢复方式
上下文文件缺失	拦截："请先运行分析器和调研工具"
两个Agent均不可用	仅执行自主可行性检查（置信度降低）
Agent超时	使用可用Agent的结果 + 自主检查结果
上下文文件格式错误	拦截："context.md缺失所需章节"

References

参考文档

shared/references/agent_review_workflow.md

— merge + debate protocol

shared/references/agent_delegation_pattern.md

— agent invocation pattern

shared/agents/prompt_templates/modes/plan_review.md

— plan review template

optimization_review_focus.md — optimization-specific focus areas

shared/references/agent_review_workflow.md

— 合并 + 研讨协议

shared/references/agent_delegation_pattern.md

— Agent调用模式

shared/agents/prompt_templates/modes/plan_review.md

— 方案审核模板

optimization_review_focus.md — 优化专项关注点

Definition of Done

完成标准

Context file loaded and validated (all required sections present)
Agent health check performed
Context materialized to
```
.agent-review/
```
for agents
Both agents launched (or SKIPPED if unavailable)
Own feasibility check completed (files exist, no conflicts, evidence backing)
Agent results merged and debated
Corrections applied to context.md
Verdict issued (GO / GO_WITH_CONCERNS / NO_GO)
Review summary saved to
```
.agent-review/review_history.md
```

Version: 1.0.0 Last Updated: 2026-03-15

上下文文件已加载并验证（所有必填章节存在）
已执行Agent健康检查
已为Agent准备好上下文并存储至
```
.agent-review/
```
已启动两个Agent（或不可用时标记为SKIPPED）
已完成自主可行性检查（文件存在性、无冲突、证据支持等）
已合并Agent反馈并完成研讨
已将修正应用到context.md
已给出结论（GO / GO_WITH_CONCERNS / NO_GO）
已将审核摘要保存至
```
.agent-review/review_history.md
```

版本： 1.0.0 最后更新： 2026-03-15