P8 Programming Expert
Paste the following prompt into your AI chat to install this skill:
Please follow the installation guide at https://skillhub.cn/install/skillhub.md to install @user_741dc82b/dev-expert.
About this skill
In software development, engineers often grapple with issues like ambiguous requirements causing rework, elusive root causes in debugging, and post-delivery bugs that slow cycles and introduce risks.
Core Workflow and Mechanisms
This skill is built on a six-step closed-loop workflow (Analysis → Plan → Execute → Verify → Deliver → Review), using a routing table to auto-route tasks to sub-skills (e.g., @code-generation, @code-review). Key capabilities include:
- Write-After-Record Protocol: Logs immediately to daily.md after substantial work like bug fixes or code generation, enabling cross-session memory retention.
- Engineering Discipline Integration: Attaches discipline layers per task type—e.g., root-cause-debugging for bug fixes mandates reproducing failures and adding regression tests; production-readiness for deployment requires six-dimension checks.
- Evidence-Based Verification: All deliveries must include verification evidence (e.g., command outputs, test reports); failures trigger backtracking to repair.
- Context Delayed Loading: Defaults to reading only entry and directly referenced files, loading historical memories or project knowledge graphs on-demand to avoid context pollution.
Application Boundaries and Considerations
Suited for complex software projects across the lifecycle, but note:
- Task-Driven Strategy: Uses intent triage (information query, simple task, complex task) to adjust autonomy; complex tasks require outputting plans first and awaiting confirmation, with batch-modification protocols for multi-module changes.
- Security Gates and Rollback: Exits fast-track for production configs or sensitive ops, demanding verification paths; releases follow progressive paths (e.g., canary), prohibiting full-blast rollouts.
- Non-Git Safety Protocol: File edits follow pre-read and local-patch principles; no delivery on verification failure.
- Mandatory Review: Post-delivery, project memories (decisions, conventions, terminology shifts) must be updated; skipping is deemed incomplete.
In summary, this skill enhances development reliability via structured workflows and discipline layers, but effectiveness hinges on strict adherence to processes, verification, and review.
Use Cases
- When needing to generate high-quality code implementations from functional requirements, use this skill for code generation with error handling and boundary condition validation.
- During code review, analyze code quality, potential security vulnerabilities, and performance issues to output a graded issue list with remediation suggestions.
- When debugging complex bugs, analyze error logs and exception stacks to pinpoint root causes, provide fixes, and ensure regression tests are added post-fix.
- During project refactoring, identify code smells, provide specific refactoring steps, and execute safe changes with verification for no regression failures.
Best For
- Backend engineers who need to quickly generate RESTful API code while ensuring compliance with security and performance standards.
- Senior developers acting as Tech Leads who must systematically review code and output improvement suggestions to enhance team code quality.
- Operations or development staff handling production incidents who need to locate bug roots, implement fixes, and add regression tests for system stability.
- Architects managing complex software projects who need to break down requirements into atomic tasks, execute based on dependencies, and ensure delivery quality.
Related Skills
Automatically indexes Gradle-cached AAR/JAR dependency classes and returns library coordinates, versions, and public APIs by fully qualified name, using only the Python standard library.
Codifies AMT and YourMT3 training conventions, script patterns, hyperparameters, precision, checkpoints, and NaN safeguards.
Retrieve relevant chunks from a customer-managed PKM dataset by dataset_id and return concise, source-annotated answers.
Convert PRDs, user stories, or functional specs into prioritized test-point checklists covering functional, business-rule, boundary, exception, and non-functional dimensions.