A professional framework for AI-driven development that prioritizes evaluation-first execution, task decomposition, and cost-aware model routing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agentic-engineering-bak
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agentic-engineering-bak using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agentic Engineering?
The Agentic Engineering skill is designed for advanced engineering workflows where AI agents handle the majority of implementation tasks while humans maintain rigorous quality and risk controls. By integrating this into Openclaw Skills, developers can move away from manual prompt-engineering toward a structured methodology that defines success criteria before execution begins. This skill emphasizes the use of specialized model tiers to balance reasoning capabilities with operational costs, ensuring that every line of code generated is both efficient and verifiable.
Agentic Engineering Use Cases
- Managing complex software implementations through agent-sized task decomposition.
- Implementing automated regression checks and capability evals for AI-generated features.
- Reducing API overhead by routing boilerplate tasks to lower-cost model tiers.
- Scaling development velocity while maintaining human oversight on security and architectural invariants.
How Agentic Engineering Works
- Establish clear completion criteria and evals before any execution starts.
- Decompose the workload into independent 15-minute units with single dominant risks.
- Route the task to the appropriate model tier based on complexity (Haiku for boilerplate, Sonnet for implementation, Opus for architecture).
- Run the implementation through an eval-first loop to compare deltas against failure signatures.
- Conduct targeted code reviews focusing on error boundaries and hidden coupling rather than stylistic choices.
Agentic Engineering Setup
To begin using Agentic Engineering within your Openclaw Skills environment, you need to configure your model routing and eval parameters.
# Initialize the agentic engineering workflow
openclaw install agentic-engineering
# Configure model routing environment variables
export CLAUDE_HAIKU_ROUTING=true
export CLAUDE_SONNET_ROUTING=true
Ensure your local development environment has automated linting and formatting enabled to streamline the AI review process.
Agentic Engineering Data Schema & Taxonomy
The skill tracks engineering performance and resource allocation using the following data structure:
| Attribute | Type | Description |
|---|---|---|
| Task Unit | String | The 15-minute verifiable unit of work |
| Model Tier | Enum | Haiku, Sonnet, or Opus selection |
| Eval Signature | Object | Baseline and failure signatures for regression checks |
| Cost Metrics | Table | Token estimates, retries, and wall-clock time per task |
Agentic Engineering Advanced Features
- Intelligent model escalation: Automatically upgrades the model tier when lower-tier reasoning gaps are detected.
- Session strategy management: Optimized session fresh-starts and compaction to maintain context window efficiency.
- Eval-first loop automation: Standardized regression checking integrated directly into the Openclaw Skills workflow.
- Cost discipline tracking: Per-task reporting on token usage and success rates to optimize project ROI.
SKILL.md
Loading
Related Openclaw Skills
Taobao Flash Sale Daily Welfare Query
A dedicated tool for querying daily hidden high-value Taobao Flash Sale coupons and red packets via API.
Attachment Inject
A high-efficiency context management skill that injects dynamic data into AI sessions as attachments rather than system prompt modifications.
PromptBuddy
PromptBuddy is an intelligent middleware skill that automatically reformats and optimizes user queries into structured prompts for superior AI reasoning.
Agent Summary
A lightweight monitoring system that provides 3-5 word real-time status updates for active sub-agents to keep workflows transparent and efficient.
7-Step Clear Explanation Framework
A structured writing methodology designed to transform complex technical concepts into accessible, high-value articles for general audiences.
AI Caricature Portrait Generator
A powerful AI tool for generating exaggerated caricature portraits and funny comic likenesses from text descriptions.
Featured*
