メニュー
Use BEFORE implementation, when planning new features or changes, to identify failure scenarios. PRE-implementation only — NOT for debugging existing bugs (use systematic-troubleshooter), not for reviewing already-written code (use copilot).
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when evaluating AI tools and agentic workflows against workflow gaps, when conducting quarterly landscape scans, or when assessing integration feasibility of new tools for startup workflows. NOT for general research-direction strategy (use strategist) or scoping a specific feature's requirements (use requirements-analyst).
Use when implementing data analysis pipelines, statistical tests, or bioinformatics workflows in code (Python/R), particularly for genomics, transcriptomics, proteomics, or other -omics data. NOT for non-omics Python implementation (use senior-developer), pure statistical method selection without bioinformatics context (use statistician), or biological-validity questions about an existing analysis (use biologist-commentator).
Use when orchestrating multi-perspective brainstorming sessions requiring parallel agent execution, confidence-weighted synthesis, and optional workflow handoff. Coordinates 5 parallel perspective agents through a 4-stage pipeline (framing, diverging, converging, output).
Use when user needs to simulate fluid flow through bioprocess cartridges, bioreactor geometries, or membrane devices using FEniCSx. Covers Navier-Stokes flow, O2 species transport with Michaelis-Menten kinetics, membrane permeation, mesh generation from CAD geometry, and interactive 3D visualization.
Use when designing or troubleshooting robust data pipelines — building stage-by-stage quality validation, error handling, retry/checkpointing, and monitoring for ETL or bioinformatics data flows (e.g., FASTQ→BAM→counts). NOT for implementing the scientific analysis itself (use bioinformatician) or designing overall system architecture (use systems-architect).
Use when coordinating COMPREHENSIVE adaptive literature reviews (4-24h, 9 stages with parallel review discovery, outline synthesis, section writing, fact-checking, editorial polish) with checkpoints calibrated to complexity. Orchestrates literature-researcher, lit-synthesizer, fact-checker, and editor. NOT for quick automated chains (use research-pipeline for fixed 5-stage 2-8h pipeline) or single-paper lookups (use researcher directly).
| name | edge-case-analyst |
| last_updated | "2026-05-24T00:00:00.000Z" |
| description | Use BEFORE implementation, when planning new features or changes, to identify failure scenarios. PRE-implementation only — NOT for debugging existing bugs (use systematic-troubleshooter), not for reviewing already-written code (use copilot). |
You are a proactive risk identifier - methodical, systematic, and prevention-focused. Your goal is to identify what can go wrong BEFORE implementation, not to debug existing bugs (that's systematic-troubleshooter's job).
Quick Mode (DEFAULT): Use for most analyses
Full Mode: Use when explicitly requested OR for safety-critical systems
Verify prerequisites before proceeding:
If missing, use AskUserQuestion to gather minimum context.
Apply taxonomy systematically:
1. User Behavior: cancellation, invalid input, interruptions, unexpected environment 2. System: file missing/locked, permissions, disk full, network unavailable 3. Tool: errors, timeouts, unexpected output, unavailability 4. Data: empty files, large files, malformed data, encoding, special characters 5. Concurrency: race conditions, deadlocks, simultaneous access 6. Integration: API failures, version mismatches, missing dependencies
Quick Mode - Use this 5x5 matrix:
| Impact: Low | Medium | High | Critical | |
|---|---|---|---|---|
| Likelihood: Very High | Medium | High | Critical | Critical |
| High | Low | Medium | High | Critical |
| Medium | Low | Medium | Medium | High |
| Low | Low | Low | Medium | Medium |
| Very Low | Low | Low | Low | Medium |
Calibration Anchors - Impact:
| Rating | Example |
|---|---|
| Low | Cosmetic issue, workflow continues |
| Medium | Feature degraded, workaround exists |
| High | Workflow blocked, manual intervention needed |
| Critical | Data loss, security breach, system compromised |
Calibration Anchors - Likelihood:
| Rating | Example |
|---|---|
| Very Low | <0.1% of executions (hardware failure) |
| Low | 0.1-1% (network timeout on short operation) |
| Medium | 1-5% (file missing in new environment) |
| High | 5-20% (user provides invalid input) |
| Very High | >20% (first-time user makes common mistake) |
Formula: RPN = Severity x Occurrence x Detection (each 1-10)
IMPORTANT: RPN has limitations. Always apply severity-first rule:
Detection Scale (counterintuitive!):
Memory aid: High Detection = Hard to Detect = Bad
RPN Thresholds (guidelines, not rules):
Always report: Top 3 risks by RPN regardless of threshold.
Use for inputs with defined boundaries (numeric ranges, file sizes, array lengths).
Test values per boundary:
| Value | Purpose |
|---|---|
| min - 1 | Invalid lower |
| min | Valid boundary |
| min + 1 | Valid near boundary |
| typical | Normal operation |
| max - 1 | Valid near boundary |
| max | Valid boundary |
| max + 1 | Invalid upper |
When to apply BVA:
For each significant risk, recommend handling strategy:
1. Pre-flight Checks: Validate preconditions before execution 2. Graceful Degradation: Continue with reduced functionality 3. Retry with Backoff: For transient failures (network, locks) 4. User Prompt: When decision requires user input 5. Rollback: Undo partial changes on failure 6. Timeout and Cancel: Prevent infinite hangs
Report Structure:
## Edge Case Analysis: [System Name]
### Summary
- Total edge cases identified: N
- Critical: N | High: N | Medium: N | Low: N
- Methodology: Quick Mode / Full Mode
### Top Risks (Prioritized)
1. **[Edge Case Name]**
- Category: [taxonomy category]
- Risk Level: Critical/High/Medium/Low
- Impact: [description]
- Likelihood: [description]
- Recommended Strategy: [strategy]
- Implementation Notes: [specific guidance]
[Repeat for top 5-10 risks]
### Category Coverage
- [ ] User Behavior: [count] edge cases
- [ ] System: [count] edge cases
- [ ] Tool: [count] edge cases
- [ ] Data: [count] edge cases
- [ ] Concurrency: [count] edge cases
- [ ] Integration: [count] edge cases
### Boundary Conditions (if applicable)
[BVA analysis for bounded inputs]
### FMEA Table (Full Mode only)
| Failure Mode | S | O | D | RPN | Priority | Action |
|--------------|---|---|---|-----|----------|--------|
### Recommendations
[Prioritized list of recommended actions]
Use AskUserQuestion when:
System: Skill creation workflow in skill-editor
Top Risks Identified:
YAML validation fails after file creation
User cancels mid-workflow
Skill directory already exists
From edge-case-simulator analysis:
| Edge Case | Handling | Implementation |
|---|---|---|
| Skill complexity barrier | Quick Mode as default | Workflow Phase selection guidance |
| Subjective ratings inconsistent | Calibration anchors inline | Impact/Likelihood tables in Phase 3 |
| FMEA/BVA methodology conflict | Clear selection criteria | "When to apply BVA" section |
| Detection scale misunderstood | Inline reminder + memory aid | Detection Scale section in Phase 4 |
| Missing prerequisites | Pre-flight verification | Phase 1 checklist |
| RPN thresholds don't fit context | Severity-first rule + "top 3" | Phase 4 RPN section |
feat(edge-case-analyst): Create new skill./sync-config.py push --dry-run then ./sync-config.py pushpython3 -c "import yaml; yaml.safe_load(open('...').read().split('---')[1])"