AI Engineering

Ultracode orchestrates multiple agents in parallel or phased pipelines for complex tasks. Use it when work can be split into independent subtasks or when you need to coordinate several approaches at once.

AI-Native Workflow routes non-trivial Playwright tasks to the right specialized skill and defines the 8-phase process for exploring, planning, gating, applying, and verifying changes. Use it when starting an AI-assisted task, choosing a skill, or coordinating a multi-step change.

Skill Creator helps create new skills, revise existing ones, and measure how well they work. Use it when you want to draft a skill, run evals, benchmark results, or improve a skill description for better triggering.

Skill Creator helps create new skills, edit existing ones, and measure how well they work. Use it when you want to draft a skill, refine its trigger or behavior, run evals, or benchmark performance.

AARM Policy Layer covers Gryph’s security policy engine, including policy evaluation, receipts, context accumulation, approvals, and identity capture. Use it when changing AARM code, policy YAML, or CLI policy wiring.

Add Agent Adapter guides the steps for wiring a new AI coding agent into Gryph. Use it when adding support for a new agent, changing adapter wiring, or updating hook detection and event parsing.

Agent Memory Implementation restructures a crowded MEMORY.md into a small pointer index plus on-demand topic files. Use it when MEMORY.md gets too long, mixes pointers with content, or needs stale memories corrected in place.

Agent Tool Builder helps define agent tools with a fail-closed pattern that keeps schema, security metadata, and execution together. Use it when creating a new tool or adding validation and permission checks to an existing one.

Aitesis Protocol scans for missing context before execution and asks prioritized questions to resolve uncertainties. Use it when an AI needs to separate factual, coherence, and relevance gaps before proceeding.

Ambiguity Detector scores project ambiguity from surface scans, optional open-question synthesis entries, and optional AST signals. It produces structured ambiguity_feed entries for the Q&A stage without making LLM calls.

Analogia Protocol validates structural mappings between abstract and concrete domains. It surfaces concrete instantiations when a mapping is uncertain and supports user grounding, repair, or withdrawal.

Approach Evaluation researches viable options for a technical or architectural decision and compares them against project constraints. Use it when multiple approaches exist and you need a factual, structured comparison without making the decision.