closedloop-ai

Generates code-grounded decision tables for behavior comparison and control-flow management, enhancing review and validation processes.

Validates plan.json files using a Python script, ensuring structural integrity and extracting relevant data efficiently.

Optimizes build processes by skipping unnecessary rebuilds when no code changes are detected, enhancing efficiency in CI/CD workflows.

Facilitates structured feedback from Codex on implementation plans, aiding in iterative development and decision-making.

Validates critic reviews by checking cache status, optimizing the critic launch process based on previous runs.

This skill assists product managers in drafting lightweight PRDs, facilitating feature definition and sprint planning through guided conversation.

Checks for cached evaluation results to optimize plan evaluation, avoiding unnecessary launches of the plan-evaluator agent.

Uploads files as ClosedLoop documents, enabling efficient management of project artifacts without consuming conversation context.

Syncs plan.md with plan.json content, ensuring updates after any modifications to maintain consistency and facilitate sharing.

Guides users in creating new skills from scratch, providing structured workflows and domain expertise.

This skill helps locate files within the Claude Code plugins cache directory, resolving paths dynamically for efficient plugin management.

Optimizes cross-repo coordination by checking cached results, reducing redundant agent launches in collaborative environments.

Enables orchestrators to refine sub-agent queries iteratively, ensuring comprehensive context for accurate outputs.

Provides conventions for editing plan.json files, ensuring structured and actionable implementation plans.

Compresses artifacts for evaluation by applying tiered summarization within a token budget, ensuring context isolation.

Orchestrates parallel execution of judge agents to evaluate implementation plans, code, and PRD artifacts, aggregating results efficiently.

Creates clear Mermaid diagrams to visualize complex concepts, system architectures, and control flows for better understanding.

Facilitates writing and parsing TOON files, optimizing data serialization for LLMs with syntax rules and examples.

Verifies and fixes high-severity code review findings, ensuring project integrity through systematic verification and correction.

Facilitates access to ClosedLoop environment paths for agents, enabling efficient file operations and resource management.