Skill UI

A governance skill designed to secure AI agent tool calls using advanced authorization policies. It implements Cedar policies to control which tools agents can access and under what conditions. Key features include running in 'Shadow mode' for safe observation, 'Enforce mode' for strict blocking, and generating cryptographically signed Ed25519 receipts to provide an immutable, tamper-evident audit trail for every policy decision.

This skill provides advanced techniques for robust prompt engineering within LangChain applications. It solves common production issues, such as f-string template breakage when user input contains literal curly braces, and guides developers on consolidating scattered prompts. Crucially, it details how to use the LangSmith Prompt Hub to version, pin, and A/B test prompts using immutable commit hashes, ensuring stable, reproducible LLM deployments.

Pattern for controlling LLM API costs while preserving quality. Combines complexity-based model routing, immutable budget tracking, retry logic limited to transient failures, and prompt caching. Use when building LLM API apps, processing batches with varying difficulty, or enforcing budget guardrails in production.

This document outlines the canonical contract for deploying model inference services on Itô compute. It details the secure handoff process of an immutable serving manifest after a compute booking is verified. The process enforces multiple strict gates, including entitlement checks, manifest integrity validation, idempotency key usage, and final health checks, ensuring robust and reliable deployment of advanced AI models.

This skill orchestrates advanced Machine Learning training workflows on dedicated compute clusters. It manages complex stages (pre-training, fine-tuning, RL) by ensuring strict adherence to immutable manifests, fresh booking entitlements, and multi-step confirmation gates. It acts as a canonical interface for job submission, emphasizing security, idempotency, and resource validation before execution.

A governance skill designed to secure AI agent tool calls using advanced authorization policies. It implements Cedar policies to control which tools agents can access and under what conditions. Key features include running in 'Shadow mode' for safe observation, 'Enforce mode' for strict blocking, and generating cryptographically signed Ed25519 receipts to provide an immutable, tamper-evident audit trail for every policy decision.