feat(skills): Add plan critic skill

.claude/settings.json

@@ -50,7 +50,8 @@
       "Skill(sentry-skills:claude-settings-audit)",
       "Skill(sentry-skills:agents-md)",
       "Skill(sentry-skills:brand-guidelines)",
-      "Skill(sentry-skills:doc-coauthoring)"
+      "Skill(sentry-skills:doc-coauthoring)",
+      "Skill(sentry-skills:plan-critic)"
     ],
     "deny": []
   },

README.md

@@ -58,6 +58,7 @@ Copy the `skills/` directory to your agent's skills location, or reference the S
 | [agents-md](plugins/sentry-skills/skills/agents-md/SKILL.md) | Maintain AGENTS.md with concise agent instructions |
 | [brand-guidelines](plugins/sentry-skills/skills/brand-guidelines/SKILL.md) | Write copy following Sentry brand guidelines |
 | [doc-coauthoring](plugins/sentry-skills/skills/doc-coauthoring/SKILL.md) | Structured workflow for co-authoring documentation, proposals, and specs |
+| [plan-critic](plugins/sentry-skills/skills/plan-critic/SKILL.md) | Critically analyze plans to identify weaknesses, gaps, and risks |
 
 ## Available Subagents
 

plugins/sentry-skills/skills/claude-settings-audit/SKILL.md

@@ -148,7 +148,8 @@ If this is a Sentry project (or sentry-skills plugin is installed), include:
   "Skill(sentry-skills:iterate-pr)",
   "Skill(sentry-skills:claude-settings-audit)",
   "Skill(sentry-skills:agents-md)",
-  "Skill(sentry-skills:brand-guidelines)"
+  "Skill(sentry-skills:brand-guidelines)",
+  "Skill(sentry-skills:plan-critic)"
 ]
 ```
 

plugins/sentry-skills/skills/plan-critic/SKILL.md

@@ -0,0 +1,85 @@
+---
+name: plan-critic
+description: Critically analyze plans to identify weaknesses, gaps, and risks. Use when reviewing implementation plans, technical proposals, project plans, or any structured plan before execution. Challenges assumptions and finds what's missing.
+---
+
+# Plan Critic
+
+Analyze a plan to find weaknesses before implementation begins.
+
+## Process
+
+### Phase 1: Understand the Plan
+
+1. Read the entire plan without judgment
+2. Identify the stated goal and success criteria
+3. List the key steps or components
+4. Note any constraints or requirements mentioned
+
+### Phase 2: Challenge Assumptions
+
+For each major element of the plan, ask:
+
+**Unstated assumptions:**
+- What is being taken for granted?
+- What prior knowledge is assumed?
+- What environmental conditions must be true?
+
+**Dependencies:**
+- What external factors must hold for this to work?
+- What could change that would invalidate this plan?
+- Are there implicit ordering requirements?
+
+### Phase 3: Find Gaps
+
+Systematically check for missing considerations:
+
+| Category | Questions to Ask |
+|----------|------------------|
+| **Edge cases** | What happens when inputs are empty, huge, malformed, or unexpected? |
+| **Failure modes** | What if a step fails? Is there rollback or recovery? |
+| **Scope** | Is anything ambiguous? Could this be interpreted differently? |
+| **Testing** | How will success be verified? What's the test strategy? |
+| **Maintenance** | Who handles this long-term? What's the operational burden? |
+| **Security** | Are there auth, access, or data exposure concerns? |
+| **Performance** | Will this scale? Are there latency or resource constraints? |
+| **Migration** | How do we get from current state to new state? |
+
+### Phase 4: Assess Risks
+
+For each identified issue:
+- **Likelihood**: How probable is this problem?
+- **Impact**: How bad if it occurs?
+- **Detectability**: Would we notice before it's too late?
+
+### Phase 5: Synthesize
+
+Produce a structured critique using the output format below.
+
+## Output Format
+
+### Summary
+[1-2 sentences on overall plan quality]
+
+### Critical Issues
+[Issues that could cause the plan to fail entirely]
+
+### Gaps to Address
+[Missing considerations that need answers]
+
+### Assumptions to Validate
+[Things the plan takes for granted that should be confirmed]
+
+### Questions for the Author
+[Specific questions that need answers before proceeding]
+
+### Suggestions
+[Concrete improvements to strengthen the plan]
+
+## Guidelines
+
+- Be specific, not vague. "What about error handling?" is weak. "Step 3 assumes the API always returns 200, but doesn't handle rate limits or timeouts" is strong.
+- Distinguish between "this is wrong" and "this is unclear"
+- Prioritize issues that would block or derail the plan
+- Don't nitpick style or formatting
+- If the plan is solid, say so - not every plan has fatal flaws