Analytics & Monitoring Agent Skills for Codex CLI

Design effective prompts for LLM agents with structured input/output formats, chain-of-thought reasoning, few-shot examples, and system…

Use when the user mentions Orizu; wants measurable improvement for a specific LLM application or agent; collects human feedback on model…

List in-flight Ouroboros sessions and show the commands needed to re-attach after MCP disconnect

How to use @owlmeans/llm-common — runtime-free serializable contracts for LLM inference and execution (ModelProvider…

Automate PagerDuty tasks via Rube MCP (Composio): manage incidents, services, schedules, escalation policies, and on-call rotations.

Optimize Palantir Foundry costs through compute tuning, incremental

Collect Palantir Foundry debug evidence for support tickets and troubleshooting.

Execute Palantir Foundry incident response with triage, mitigation,

Configure Palantir Foundry local development with Python transforms

Optimize Palantir Foundry API performance with caching, batching, and

Upgrade Palantir Foundry SDK versions and handle breaking changes.

Expert-thinking profile for Paleontologist (field / museum / computational): Reason from taphonomic filters first: biostratinomy…

Performs pandas DataFrame operations for data analysis, manipulation, and transformation.

Use when verifying a scientific claim or numerical parameter through project artifacts, implementation semantics, primary literature…

Analyzes research papers (PDF/arXiv URL) and converts them into executable code.

Load a sharded, on-disk dataset (sharded .npy, Parquet/Arrow, raw binary, sharded HDF5, custom layouts) into a distributed cuPyNumeric…

Audits architectural patterns against best practices (MCP Ref, Context7, WebSearch).

Creates a DAX measure from a description in Portuguese, respecting the existing Power BI model (real columns, naming convention, DAX…

Guidance for Power BI Project (PBIP) structure, thick and thin reports, project renames, forks, and validation.

Per-category/per-family models can produce WORSE leaderboard scores than a single global model, even when per-category CV is better.

Profile and optimize hotspots.