This skill should be used when the user runs "/commandments", asks to "evaluate the codebase against engineering commandments", "assess engineering maturity", "run a commandments assessment", "check codebase maturity levels", "how mature is this codebase", "engineering best practices audit", "code quality assessment against commandments", or "generate a commandments report". It evaluates the current repository against 10 Engineering Commandments using a 5-level maturity model and generates a detailed report with evidence-backed assessments and stack-specific actionable improvements.
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
This skill should be used when the user runs "/commandments", asks to "evaluate the codebase against engineering commandments", "assess engineering maturity", "run a commandments assessment", "check codebase maturity levels", "how mature is this codebase", "engineering best practices audit", "code quality assessment against commandments", or "generate a commandments report". It evaluates the current repository against 10 Engineering Commandments using a 5-level maturity model and generates a detailed report with evidence-backed assessments and stack-specific actionable improvements.
Evaluate the current repository against 10 Engineering Commandments, producing an evidence-backed maturity report with researched, prioritized actionable improvements.
Pipeline Overview
Execute these phases sequentially, providing progress updates after each:
Stack Detection -- Identify the repository's tech stack
Stack Research -- WebSearch for stack-specific best practices
Deep Evaluation -- Assess each commandment with evidence
If no manifests found, analyze file extensions and directory structure as fallback. Clearly note "stack detected via heuristics" in the report.
Phase 2: Stack Research
Using ONLY technology names (never internal code details), run WebSearch queries:
"[Framework] error handling best practices [current year]"
"[Framework] testing best practices [current year]"
"[Framework] observability logging monitoring [current year]"
"[Framework] security best practices OWASP [current year]"
Limit to 3-5 targeted searches for the primary stack components. Store findings for use in actionable recommendations.
Phase 3: Deep Evaluation
For EACH of the 10 commandments, use the Agent tool (subagent_type: Explore) to perform deep codebase analysis. Launch agents in parallel clusters of 3-4 for efficiency.
Locate the bundled commandments reference for the full 5-level maturity criteria:
Glob: **/commandments.md
Read this file to load the maturity level definitions before assessing each commandment.
Evidence Patterns Per Commandment
For each commandment, search for these specific evidence patterns:
Check: Load balancer or horizontal scaling configuration
Maturity Level Assignment
For each commandment, compare evidence found against the 5-level criteria in commandments.md. Assign the HIGHEST level where ALL criteria are substantially met. When in doubt, assign the lower level and note what's missing.
Each assessment MUST include:
Level assigned (1-5)
Evidence found (specific file paths, patterns, counts)
Evidence missing (what would be needed for the next level)
Phase 4: Report Generation
Report Location
Write to claudedocs/commandments-report.md. Create the claudedocs/ directory if it doesn't exist.
Existing Report Handling
If the report already exists:
Read it to extract the ## Assessment History section
Generate the new report with updated assessments
Append the previous assessment summary to the history section
Report Template
# Engineering Commandments Assessment Report**Repository:** [repo name from directory]
**Date:** [YYYY-MM-DD]
**Overall Maturity:** [average level, e.g., "Level 2.4 / 5"]
## Tech Stack
| Category | Technologies |
|----------|-------------|
| Languages | [detected] |
| Frameworks | [detected] |
| Testing | [detected] |
| CI/CD | [detected] |
| Infrastructure | [detected] |
## Maturity Summary
| # | Commandment | Level | Score |
|---|------------|-------|-------|
| 1 | Design for Failure | Level X | X/5 |
| 2 | Keep It Simple | Level X | X/5 |
| ... | ... | ... | ... |
| | **Overall** | | **X.X/5** |
## Detailed Assessments### 1. Design for Failure - Level X/5**Evidence Found:**- [specific file paths and patterns found]
- [counts and metrics]
**Evidence Missing (for next level):**- [what would be needed to reach Level X+1]
**Assessment Rationale:**
[Brief explanation of why this level was assigned]
[Repeat for all 10 commandments]
## Actionable Improvements (Prioritized by Impact)### Priority 1: [Highest impact action]**Commandment:** [which one]
**Current Level:** X -> **Target Level:** Y
**Effort:** Low/Medium/High
**Impact:** [why this matters most]
**Steps:**1. [Specific, stack-aware step]
2. [Reference to best practice from WebSearch]
[Repeat for top 5-10 actions]
## Assessment History
| Date | Overall Score | Top Improvement | Notes |
|------|--------------|-----------------|-------|
| [previous dates and scores from existing report] |
| [current date] | [current score] | [biggest change] | [notes] |
Actionable Recommendations
For each recommendation:
Reference the SPECIFIC tech stack (e.g., "Add Express error-handling middleware" not "Add error handling")
Include code patterns or library names from WebSearch findings
Estimate effort (Low/Medium/High)
Order by: impact score = (level gap x commandment criticality) / effort
Label any recommendation that falls back to general best practices (for unknown stacks) with: [General Best Practice]
Flags
--quick: Skip WebSearch research phase, use only codebase analysis
--focus=NAME: Deep-dive into a single commandment (e.g., --focus=security)
