Mark Wang
Community@wangyouan · Xiamen
Finance PhD
Agent Skills by Mark Wang
Showing 22 vetted skills indexed across 1 GitHub repositories.
wangyouanmulti-agent-patterns
Design multi-agent systems with coordination protocols, consensus mechanisms, and failure handling.
wangyouanbdi-mental-states
Construct BDI mental states from RDF data for reasoning and decision-making.
wangyouancontext-degradation
Diagnose context degradation and attention distribution in LLM agent systems.
wangyouancontext-compression
Compress large token contexts to minimize information loss in agent sessions.
wangyouanmemory-systems
Persist agent memory across sessions using vector stores and knowledge graphs.
wangyouanadvanced-evaluation
Evaluate LLM outputs with direct scoring, pairwise comparison, and bias mitigation.
wangyouanharness-engineering
Design control systems for autonomous agents with feedback and persistence mechanisms.
wangyouancontext-fundamentals
Explain context engineering principles for AI agents.
wangyouanguide-updater
Update project guidelines incrementally during AI interactions using Python scripting.
wangyouanlatent-briefing
Share relevant orchestrator trajectory parts with workers via Attention Matching.
wangyouanproject-development
Assess LLM task suitability and design staged pipeline architectures.
wangyouanevaluation
Evaluate agent system performance using deterministic checks and multi-dimensional rubrics.
wangyouanjoplin-notes
Manage local Joplin notes via the Joplin Web Clipper API.
wangyouanget-review-theme
Extract structured review topics from files, images, URLs, and descriptions.
wangyouancontext-optimization
Apply context budgeting, observation masking, KV-cache optimization, and context partitioning to AI systems.
wangyouanhumanizer
Rewrite AI-generated text into natural, human-like prose.
wangyouanxmu-mail
Access and manage @xmu.edu.cn email via IMAP and SMTP with SSL.
wangyouanstata-regression-workflow
Automate Stata regression workflows with log inspection and CSV/XLSX result reporting.
wangyouanfilesystem-context
Store and retrieve agent context using the filesystem as a persistent overflow layer.
wangyouantool-design
Design and optimize tool interfaces for agent systems.
wangyouanhosted-agents
Design scalable hosted agent infrastructures with sandboxed execution and session persistence.
wangyouancheck-review-alignment
Check reference accuracy in review documents using AI semantic alignment.