Optimize or generate Claude resources using the CGF pipeline.
npx skills add https://github.com/andisab/casdk-harness --skill cgf-optimize
Please help me install this Agent Skill. Skill: cgf-optimize Source: https://github.com/andisab/casdk-harness/tree/main/src/harness/plugins/cgf-agents/skills/cgf-optimize Command: npx skills add https://github.com/andisab/casdk-harness --skill cgf-optimize
This Skill solves the problem of manual and inefficient optimization and creation of Claude resources by leveraging CGF (Claude Gradient Feedback), making it possible to enhance existing resources or generate new ones with precision.
/cgf-optimize command to enhance an existing agent's ability to handle async programming patterns or create a new Kubernetes deployment agent with a review process.To optimize an existing Python expert agent for handling async programming, use the command:
/cgf-optimize python-expert async programmingHigh-intent search queries and answers about installing and using this skill.
Claude agent optimization for specific patterns uses the CGF pipeline to refine resources via self-critique and domain best practices. You provide the existing resource name and target enhancement, such as async programming, to generate an improved agent.
New Claude resource creation from descriptions initiates the CGF optimization pipeline to build assets like a Kubernetes deployment agent. You specify the desired resource functionality and the process investigates domain best practices to generate it.
CGF, or Claude Gradient Feedback, is a context-engineering technique that refines resources through self-critique. You need it when manual resource optimization becomes inefficient or when generating new automated development assets with precision.
The CGF optimization process requires no external dependencies, operating solely through scripts. It does require input specification detailing either the existing resource to enhance or the parameters for the new resource to create.
Streamlining context-engineering for automated development is best achieved by leveraging the CGF pipeline to replace manual refinement. It enhances or generates Claude resources by utilizing self-critique and investigating domain best practices.
Refining Claude commands with CGF requires clear input specifications for the optimization pipeline to function. Without precise existing or new resource definitions, the self-critique and domain best practices investigation cannot effectively execute.
dlt-hubAdd a REST API endpoint to an existing dlt pipeline configuration.
NourcineAbConfigure Claude Code hooks for validation, formatting, and session persistence.
cameronraysmithScrape and aggregate Anthropic prompt engineering documentation using Firecrawl.
hitoshisekiCreate standardized Claude skills with validation and packaging scripts.