HengJun Wang

Enhancement-overlay skill — the DECISION + WIRING layer for LM observability that the single-backend skills [[langsmith]], [[phoenix]], [[mlflow]] do NOT cover. Each of those installs one backend; none of them help you DECIDE which backend fits your stack/scale/budget, nor give you a one-line autolog that turns it on fast. Use when starting any LM project, before the first deploy, or the moment someone asks "why did it do that?" and there are no traces to answer with. The skill picks a backend by stack (LangSmith for LangChain/LangGraph; Phoenix for OSS/local OpenTelemetry; MLflow for ML-shops already on MLflow; Langfuse for self-host), wires one-line autolog, verifies traces land, and adds eval hooks — instrumenting BEFORE you need it. Cross-links the first-debug-move skill [[agentsop-prompt-history-inspect]]. Do NOT activate to re-teach a backend you already chose (defer to its own skill), or for non-LM ML experiment tracking with no LLM calls (that is plain MLflow).

The English summary is being prepared. Showing the original description.

Pick an LM output format per (task x consumer x model) rather than by reflex: different formats carry different cognitive load (e.g. code-in-JSON makes the same model write worse code than plain-text+diff, while asking for prose when you need a typed object fails the other way). Use when designing or debugging an LM's output schema, choosing between plain text / diff / JSON / tool-call / grammar-constrained output, or when a model's quality drops after wrapping its output in a structured format.

The English summary is being prepared. Showing the original description.

Lifecycle SOP for **per-model prompt artifacts** — the compiled prompts, instructions, few-shot demos, edit-format pins, and embedding-bound indices that change behavior when the underlying LM, dataset, or framework version changes. Activate when adopting compiled prompts (DSPy, GEPA, BootstrapFewShot output), when supporting multiple LMs in production, when a provider deprecates a model snapshot, or when a framework deprecates a config surface (LlamaIndex `ServiceContext` → `Settings`, Aider edit-format defaults). Do NOT activate for one-off raw prompt edits or for truly model-agnostic system prompts that have been swap-tested. Search keywords: prompt portability, model swap, recompile prompt, model deprecation, prompt per model, prompt breaks on new model, version compiled prompts.

The English summary is being prepared. Showing the original description.

The compile-readiness gate for prompt auto-optimization. Decide whether you have earned the right to run an optimizer (DSPy MIPROv2 / GEPA / BootstrapFewShot) before spending compute. Two preconditions only — a real metric, and enough examples for the optimizer you picked. Garbage metric in, garbage prompt out. Pick the optimizer by data scale; GEPA inverts the scale assumption (~10 examples + textual feedback).

The English summary is being prepared. Showing the original description.

Tool skill — the *first move* in any LM-debugging session: dump the actual rendered prompt the framework sent to the model, before changing anything else. Activate when an LM call produced an unexpected output (wrong answer, schema violation, refusal, truncation, cost spike, latency spike, infinite loop, "model got dumber after upgrade"). The skill enforces a 30-second inspect step BEFORE any prompt edit, model swap, retry, or temperature tweak. Cross-framework cheat sheet: DSPy `inspect_history`, LangGraph `get_state_history`, CrewAI `step_callback`, LangChain `set_debug`/`set_verbose`, Aider `/diff`+`--verbose`, raw OpenAI/Anthropic via `OPENAI_LOG=debug`/`ANTHROPIC_LOG=debug` or HTTPX event hooks. Do NOT activate for first-time prompt authoring, exploratory prompt design, or non-LM bugs.

The English summary is being prepared. Showing the original description.

Routes heterogeneous queries to the appropriate index, tool, or answering engine before retrieval. Use when one endpoint serves multiple handlers, such as summary, vector retrieval, text-to-SQL, or tools, and query types require different paths. Covers LLM, semantic, and rule-based routers, confidence thresholds, fallbacks, and framework mappings. Do not use when one handler serves all queries or branching is fixed.

The English summary is being prepared. Showing the original description.

A 5-minute gate the coder runs at project kickoff (and again whenever the repo shape changes). Classifies the workspace into Greenfield / Brownfield-large / Mid-size-familiar / Library-SDK, then maps the state to an agent strategy (autonomy, context primitive, tool choice). Use BEFORE picking Cursor vs Claude Code vs Aider, BEFORE turning on repo-map, BEFORE writing the first prompt. Skip only if the same repo was gated within the last day and nothing changed. Search keywords: greenfield vs brownfield, new project vs existing codebase, project setup strategy, legacy codebase agent, where to start a coding agent.

The English summary is being prepared. Showing the original description.

Adds and tunes a reranker stage for RAG using the retrieve-wide, rerank-narrow pattern. Use when relevant documents appear in the initial top-N but are buried by noise, top-1 precision or MRR is low despite adequate recall, or too many marginal chunks consume context. Covers cross-encoder, API, and local rerankers; N-to-k selection; and latency/cost tradeoffs. Do not use when retrieval recall itself is failing.

The English summary is being prepared. Showing the original description.

Project-kickoff rubric for the self-host vs managed-cloud decision — when is running your own inference engine / LLM platform worth the ops cost vs paying per-token for a managed API? Decide on two axes — VOLUME (a cost-crossover slider) and COMPLIANCE (a hard gate). Use at kickoff when choosing where to run inference, or when cost / data-residency pressure forces a re-evaluation.

The English summary is being prepared. Showing the original description.

Decision protocol for managing the context/session state of an AI coding tool: when to /clear, when to keep context, and how to detect "context bleed" — the failure mode where stale conversation history biases the model against the current task. Surfaces a discipline that Aider (/clear), Claude Code (/clear), CrewAI (memory=False, re-instantiate), and LangGraph (new thread_id, subgraph isolation) all encode separately but none name as a skill.

The English summary is being prepared. Showing the original description.

Decision rubric for promoting a prose prompt into a typed DSPy Signature. This is an ENHANCE overlay on top of the [[dspy]] library skill: it does NOT teach DSPy syntax — it answers the coder-agent decision "when do I stop hand-writing a prompt string and declare it as a `dspy.Signature`, and how do I name/describe its fields so the optimizer and the calling code both get a clean contract." Activate when: a prompt string grows past ~50 lines; the LM output is consumed by code (parsed, branched on, stored) rather than read by a human; the same prompt is reused across >1 call site; or a teammate asks "should this be a Signature?". Do NOT activate for one-shot throwaway prompts, or for HOW-TO questions about DSPy modules /optimizers/compile — defer those to the [[dspy]] skill and the [[agentsop-dspy]] workflow skill. Search keywords: typed prompt, structured prompt, DSPy Signature, prompt as a function, prompt contract, when to formalize a prompt.

The English summary is being prepared. Showing the original description.

Tool skill for declaring reducers on LangGraph state keys so parallel writes merge instead of crashing. Activates whenever a coder agent designs a StateGraph with parallel branches, fan-out via Send, multi-agent topologies, or whenever a run raises `InvalidUpdateError: At key '<k>': Can receive only one value per step`. Encodes the rule "every state key is single-writer or has a reducer — nothing in between."

The English summary is being prepared. Showing the original description.

Enhancement-overlay decision protocol for STREAMING the output of long-running LLM / agent runs from the *backend*, not just wiring a typing animation in the UI. Activates when a coder agent must stream final tokens to a chat client, surface intermediate agent steps (which tool, which node, partial reasoning), emit custom tool-progress events, choose a transport (SSE vs WebSocket), or decide what to do when the client disconnects mid-stream. The langchain / langgraph skills mention stream modes but stop at "you can stream"; this skill encodes *what to stream, over what transport, and how to fail safely*.

The English summary is being prepared. Showing the original description.

Decide where to enforce structured LM output (constrain at decode time with Outlines vs validate-and-retry with Instructor vs grammar with Guidance) and which failure stance to take (Assert/hard-fail vs Suggest/soft-retry). Use when an LM's output is parsed or typed by downstream code and you must pick one enforcement library plus its failure handling, when malformed output is burning tokens on retries, or when choosing between decode-time vs validation-time constraints for local vs API models.

The English summary is being prepared. Showing the original description.

Decision protocol for wiring a verify-then-fix loop around a code-editing LLM agent. The agent edits → runs lint/test → reads the output → fixes → re-runs, bounded by an iteration cap and an escalation rule. Activates whenever a coder agent has a verifiable success criterion (exit code, type-checker output, failing assertion) and the user wants the agent to converge to "green" on its own. Framework-agnostic — wraps Aider's `--auto-lint`/`--auto-test`, an OpenHands SWE-Bench loop, a manual LangGraph cycle, or Claude Code's bash tool just the same.

The English summary is being prepared. Showing the original description.

Enhancement overlay for multi-agent / tool-using coder agents. Encodes the per-agent tool- scoping discipline that role-based frameworks (CrewAI, LangChain) document only as a passing best-practice: which agent gets which tool, and why blanket-sharing every tool to every agent is a correctness and blast-radius risk. Activates when an agent system has tools AND there is more than one agent (or one agent holding many tools). Treat a tool as a capability grant; scope by least-privilege. ENHANCE overlay — read alongside [[crewai]], [[agentsop-http-tool-wrapping]], [[agentsop-llm-tool-idempotency]]. Search keywords: which tools per agent, least-privilege agent, agent tool access, tool permissions, limit agent tools, scope tools to roles.overlay_type: enhancement

The English summary is being prepared. Showing the original description.

Decision SOP for serving LLMs with vLLM. Covers PagedAttention mental model, quantization/parallelism/batching tradeoffs, OOM triage, and when NOT to use vLLM. Activates when a coder-agent is choosing or tuning an inference engine, debugging vLLM throughput/latency/OOM, or comparing vLLM against TGI/SGLang/TensorRT-LLM/llama.cpp.

The English summary is being prepared. Showing the original description.

LEAP builds skills through two pipelines: Branch A distills a skill from raw data, while Branch B combines multiple skills into one. It is called by the main SkillAlchemy workflow. Use when SkillAlchemy requires distillation or fusion.

The English summary is being prepared. Showing the original description.

Lens — Add a cognitive lens to any problem. It accepts a task description and produces an enhanced description that surfaces hidden dimensions, prerequisites, and lines of inquiry—the things you do not know you do not know. Use when the user asks to brainstorm, analyze, generate a skill, distill, or fuse, or when the input is too simple and needs to be expanded.

The English summary is being prepared. Showing the original description.

SkillAlchemy — One thought conceived, one goal achieved. Accept any idea or distillation target and produce an installable SKILL.md. It uses Lens to clarify the problem and LEAP to run distillation or fusion. This is the sole user-facing entry point. Use when the user asks to distill, generate a skill, fuse skills, or says, "I want to build X, but I do not know where to start."

The English summary is being prepared. Showing the original description.

Decision rubric for when an LM agent should write-and-run code (Program-of-Thought / code interpreter) versus reason in natural language: classify each step as deterministic- computable (emit + execute code, feed the result back) vs judgment (stay in prose). Use when designing or debugging an agent step that does arithmetic/parsing/data transforms, when prose reasoning hallucinates a computation (under-coding), or when a sandbox round- trip is wasted on a judgment task (over-coding). Search keywords: code interpreter, agent does math wrong, calculator hallucination, when to run code vs reason, program of thought, PoT, tool vs reasoning.

The English summary is being prepared. Showing the original description.

Operating-system distillation of LlamaIndex — the leading RAG / document-agent framework. Activate when the calling agent must build, debug, harden, or evaluate a Retrieval-Augmented Generation pipeline over unstructured/private data, decide between RAG primitives (Index types, retrievers, query engines, routers, agents), or pick LlamaIndex vs LangChain / Haystack / raw vector store for a coding task. Encodes the 5-layer mental model (Documents → Nodes → Indices → Retrievers → Query Engines / Response Synthesizers), the canonical RAG bootstrap SOP from baseline `VectorStoreIndex` through hybrid + reranker + eval-loop hardening, the official 13-failure-mode checklist, and 5 dilemma cases distilled from docs, GitHub issues, and 2025 production post-mortems.

The English summary is being prepared. Showing the original description.

Split a multi-call LM workflow by cognitive load, not by accuracy: let one strong model make the few reasoning decisions and a cheap model do the many mechanical executions (Aider architect+editor, DSPy optimizer-LM vs task-LM, vLLM speculative draft+target, LangGraph supervisor+worker are the same shape). Use when designing or cost-optimizing a pipeline that calls an LM many times, when deciding which steps need a strong reasoner vs a cheap executor, or when adding an escalation valve for when the cheap tier degrades. Search keywords: reduce LLM cost, cheaper model, lower token cost, model cascade, route to cheap model, strong model plus cheap model, LLM cost optimization.

The English summary is being prepared. Showing the original description.

Re-ingest-correctness SOP for production RAG. Activate when a calling agent builds, reviews, or debugs an ingestion pipeline that runs more than once over a changing corpus — scheduled re-index, incremental updates, CI re-ingest, or a "retrieval has duplicates / shows deleted docs" bug. Encodes the rule — **ingestion must be idempotent: a document's content hash decides insert/update/skip, so re-running over unchanged docs is a no-op** — plus the docstore + doc-hash upsert machinery (LlamaIndex `IngestionPipeline` + `DocstoreStrategy`), the delete-propagation problem, and cross-framework equivalents (LangChain `index()` + `RecordManager`, manual hash table). ENHANCE overlay over [[llamaindex]]: the IngestionPipeline exists in the base skill but the re-ingest-correctness contract is not surfaced.

The English summary is being prepared. Showing the original description.