Route user intents to LDD sub-skills via trigger-phrase mappings.
npx skills add https://github.com/veegee82/loss-driven-development --skill using-ldd
Please help me install this Agent Skill. Skill: using-ldd Source: https://github.com/veegee82/loss-driven-development/tree/main/skills/using-ldd Command: npx skills add https://github.com/veegee82/loss-driven-development --skill using-ldd
The LDD bundle conductor helps you dispatch the twelve LDD skills and their trigger phrases, acting as the entry-point that routes user intents to the correct sub-skills and tracks the trigger-phrase table they respond to.
Prefix your message with LDD: to activate the orchestrator and dispatch the appropriate sub-skill.
High-intent search queries and answers about installing and using this skill.
Orchestrating agent workflows with gradient descent thinking levels involves dispatching sub-skills across inner, refinement, outer, and CoT loops. This Skill routes user intents to the correct sub-skills and applies level scoring to optimize the agent's processing logic.
LDD trigger-phrase routing is a mechanism that maps specific keywords to designated agent workflows. It maintains a mapping table so that mentions of LDD or related terms automatically dispatch the correct sub-skill, guiding the agent through four axes of thinking.
To auto-dispatch sub-skills across multiple thinking loops, prefix your message with LDD: to activate the orchestrator. It then selects the appropriate sub-skill from L0–L4 based on user signals and tracks decisions through a visible trace interface.
You can trace agent decisions and progress across different thinking levels using the built-in trace interface. This interface surfaces the orchestrator's routing choices, level scoring, and workflow progress as the agent navigates the four parameter spaces.
No specific dependencies are required to run an LDD orchestrator for agent workflows. The Skill operates independently to coordinate the twelve LDD skills, applying trigger-phrase mappings and gradient-descent logic without external component installations.
The best way to manage memory and governance across agent thinking loops is through a centralized orchestrator entry point. This approach reuses shared memory and governance mechanisms across the four parameter spaces, ensuring consistent state throughout the workflow.
jeffdevilleAutomate skill activation when users mention trigger phrases in conversation.
mbadozRefine codebase context for subagents through a four-phase retrieval loop.
paulinghamCoordinate capture, run, score, and validate sub-skills into an end-to-end evaluation workflow.
truchotCoordinates governance, resources, and prioritization across projects via a multi-domain routing framework.
