mcp-ai-agent-guidelines
Version:
A comprehensive Model Context Protocol server providing advanced tools, resources, and prompts for implementing AI agent best practices
29 lines • 1.15 kB
TypeScript
/**
* Prompt Chaining Builder - Creates sequential/parallel multi-step AI workflows
*
* Inspired by claude-flow's orchestration patterns, this tool enables:
* - Progressive refinement through multi-step prompting
* - Data flow between steps via output keys and dependencies
* - Flexible error handling (skip, retry, abort)
* - Automatic visualization of chain structure
*
* Integration with Serena patterns:
* - Context-aware execution (uses project memories)
* - Planning-first approach (design chain before execution)
* - Memory optimization (chains can be cached for reuse)
*
* Common use cases:
* - Security analysis pipelines (scan → assess → remediate)
* - Code review workflows (static → performance → recommendations)
* - Documentation generation (outline → draft → refine)
*
* @param args - Chain configuration (validated against PromptChainingSchema)
* @returns Formatted chain specification with visualization
*/
export declare function promptChainingBuilder(args: unknown): Promise<{
content: {
type: string;
text: string;
}[];
}>;
//# sourceMappingURL=prompt-chaining-builder.d.ts.map