HengJun Wang
Community@agentsope · dayuncun：）
BUAA major in CS
Agent Skills by HengJun Wang
Showing 52 vetted skills indexed across 2 GitHub repositories.
agentsopefengge-capital
Provide humorous and irreverent insights on stock market topics.
agentsopefengge-depressed
Provide emotional support and humorous perspective on life's challenges.
agentsopefengge-anime
Simulate a humorous 48-year-old perspective on anime, Vtuber, and ACG topics.
agentsopefengge-drunk
Generate humorous emotional responses in a drunk persona style from prompts.
agentsopeSkillAlchemy
Distill and fuse source materials into installable SKILL.md files.
agentsopeagentsop-code-execution-decision
Route agent steps to code execution or prose reasoning via a deterministic gate.
agentsopeagentsop-state-reducer
Resolve LangGraph InvalidUpdateError by defining per-key reducer semantics.
agentsopeagentsop-context-scope-discipline
Enforce a read-vs-edit split and ~25k token budget for coding agents.
agentsopeagentsop-metric-design
Build decomposed, bias-mitigated evaluation metrics for LLM pipelines.
agentsopeagentsop-query-routing
Classify queries and dispatch them to the correct index, tool, or engine before retrieval.
agentsopeagentsop-observability-setup
Select a tracing backend and wire one-line autologging for LLM projects.
agentsopeagentsop-aider
Guide Aider terminal sessions for git-native code editing with validation loops.
agentsopeagentsop-llm-tool-idempotency
Implement idempotency keys to prevent duplicate side-effectful tool executions.
agentsopeagentsop-llm-engine-selection
Recommend LLM inference engines using hardware, workload, and constraint vectors.
agentsopeagentsop-streaming-output
Implement a backend streaming protocol with SSE or WebSocket multiplexing and disconnect policies.
agentsopeagentsop-dify
Convert Dify LLM application workflows into structured SOP decision patterns.
agentsopeagentsop-cost-tiered-models
Split LLM pipeline calls into strong decision and cheap execution tiers.
agentsopeagentsop-conventions-pinning
Write and pin project-level conventions files for multi-tool coder-agent environments.
agentsopeagentsop-crewai
Guide CrewAI multi-agent orchestration decisions with task contracts and safeguards.
agentsopeagentsop-regression-gate
Enforce LLM regression testing in CI with a held-out eval set and noise-aware threshold.
agentsopeagentsop-session-state-hygiene
Detect and remediate context bleed in multi-turn AI coding sessions.
agentsopeagentsop-map-reduce-fanout
Design map-reduce fan-out protocols for parallel LLM pipelines with bounded concurrency and timeouts.
agentsopeagentsop-prompt-history-inspect
Inspect rendered prompts sent to language models across DSPy, LangChain, and OpenAI SDKs.
agentsopeagentsop-llamaindex
Build, debug, and evaluate LlamaIndex RAG systems for grounded document answers.