AI Agent & Automation Workflows

Select and apply the right agentic design pattern when building an AI agent, workflow, or multi-step LLM system. Use this skill whenever the user is designing, architecting, debugging, or reviewing an agent/LLM pipeline and needs to decide HOW to structure it -- e.g. they mention agents, orchestration, multi-step workflows, RAG, routing, tool use, retries, evaluation, guardrails, memory, planning, or ask "how should I build/structure this agent?" or "which pattern fits this problem?". Maps problem symptoms to one of 20 patterns and specifies, for each, which steps are LLM calls vs deterministic code and what context each call carries. Use it even when the user does not say the word "pattern".

All-in-one LLM CLI tool for sending prompts, switching models, piping stdin, including files, and using roles/sessions. Use as an AI adapter for shell workflows.

Use this skill for analytics and data-science workflow setup, exploratory analysis, notebook-first EDA, repo normalization for analysis projects, experiment comparison, AutoML, causal analysis, and promotion from ad hoc exploration into reusable pipelines. Trigger when the user asks for analysis best practices, how to structure an analytics repo, how to organize notebooks and runs, whether to use marimo or Quarto/qmd, how to handle experiment sweeps, how to compare models, or how to make analysis reproducible. Also trigger on phrases such as analytic workbench, EDA, exploratory analysis, notebook workflow, analytics pipeline, reproducible analysis, experiment sweep, hyperparameter comparison, comparison table, marimo, Quarto, qmd, Hamilton, sf-hamilton, dataflow, DAG driver, Hydra, DVC, Kedro, MLflow, AutoML, PyCaret, causal analysis, feature engineering, or model review.

Instrument agentic LLM apps built on the Claude Agent SDK (claude-agent-sdk) and/or LangGraph with Arize Phoenix and OpenInference — tracing, evaluation, annotations, experiments, cost tracking, and self-hosting. Use when the user mentions Phoenix, arize-phoenix, openinference, LLM observability, LLM-as-judge evals, tracing Claude Agent SDK query() / ClaudeSDKClient calls, tool-use observability, tracing LangGraph nodes/edges, or debugging latency/cost/quality of an agent.

Build MCP servers, author routes/pipelines/functions, and manage all Cribl Stream artifacts (Stream, Edge, Search, Lake) via REST API and YAML configs. Use when user mentions Cribl, Cribl Stream/Edge/Search/Lake, routes, pipelines, sources, destinations, packs, or observability pipeline management.

Convert UI designs into structured JSONC spec packages before code is written, especially for constrained platforms like extensions, dashboards, desktop shells, and mobile apps. Use for design handoff and design-to-spec workflows. Outputs specs, not implementation code.

Generate, modify, or debug Kestra Flow YAML by fetching the live flow schema and applying the same guardrails used by the Kestra AI Copilot. Use when users ask to create, write, update, or fix a Kestra flow.

Operate Kestra environments using kestractl for context setup, flow inspection, flow validation and deployment, execution monitoring, namespace operations, and namespace file management. Use when users request Kestra operational CLI tasks in dev, staging, or production.

Use this skill for LangGraph, Deep Agents, LangChain agents built on LangGraph, MCP-to-LangGraph tool bridging, stateful workflows, subgraphs, subagents, interrupts, checkpointing, streaming, and multi-agent orchestration. Trigger when code imports langgraph, deepagents, langchainmcpadapters, langchain.agents, or when the user asks for agent graphs, orchestration, durable execution, HITL, or LangGraph architecture and patterns.

Recover a prior Claude Code session from natural-language hints, search Claude history by topic/date/project, and import the useful context into the current conversation. Use this for Claude session handoff, transcript recovery, context transfer into Codex or another agent, and continuing after Claude hit a usage or rate limit.

Use this skill for general Robot Framework work: authoring .robot suites, tasks, keywords, variables, resource files, execution, dry runs, tags, Rebot/Libdoc usage, and Python test-library patterns. Trigger when the user mentions Robot Framework, .robot files, keywords, libraries, resource files, tasks, listeners, Libdoc, Rebot, or Robot Framework syntax and execution.

Deep skill for Splunk development, administration, SDK/REST integrations, dashboards, UCC add-ons, ITSI automation, SPL2 authoring, and AI-facing tooling. Use for Splunk SDK, REST, jobs/export, SPL, dashboards, packaging, and MCP-backed analysis workflows.

Use when searching a codebase by meaning or exact text — fast hybrid (BM25 + static-embedding semantic) local code search for agents and humans, plus tree-sitter symbol navigation. Use for conceptual queries ("where is auth handled", "retry logic"), finding the right repo/file from a description, find-related code, exact symbol lookup (defs/refs/outline), and a single-binary stdio MCP / gRPC server. Pure-Rust single binary (ships anywhere, incl. musl/aarch64), CPU-only, offline. Replaces ck and semble for local search. Triggers include "search the code for", "find where X happens", "which file does Y", "semantic code search", "find similar code", "find definition/references of". The fork build adds --watch (live incremental index) and a --dashboard web UI. For PDF/prose document search use semtools.

Design and implement distinctive, production-grade screen UI for vibe-coding workflows: loose prompts, taste-driven iteration, and fast frontend polishing that still respects usability, layout discipline, and real code constraints. Use this skill whenever the user wants a frontend to feel more polished, modern, premium, cohesive, bold, or simply "better", including landing pages, dashboards, app screens, forms, lists, onboarding, settings, and component restyling. Also use it for vague requests like "make this UI nicer", "give this more personality", "make it feel expensive", or "vibe code this screen." Preserve existing design systems when they are good; introduce a clearer visual and interaction system when they are weak or absent.

Structured browser extraction for AI coders — explore first, then draft repeatable Robot Framework BDD suites with shipped generic keywords, templates, and validation harness.

Structured web scraping for AI coders: explore, then exploit with shipped templates, runner, and hooks.

Umbrella skill for agent work discipline across development, analysis, and documentation: inspect the repo before restructuring, keep durable truth in repo artifacts instead of chat memory, co-evolve specs/design/steering/user docs with code, apply sound coding patterns, verify work honestly, avoid shortcuts, work efficiently with subagents without hallucinating, and keep moving through the next concrete work item when the human is away. References cover coding patterns, AI-authored code review, and artifact co-evolution. Trigger when the user asks for workflow discipline, coding patterns, doc/artifact maintenance, code review of AI-authored code, project hygiene, execution guardrails, repo normalization, or when a task risks drifting across architecture, storage, specs, continuity, or tooling boundaries.