From 9df5c53bb5ebd590c24861bebfcfcfdd4a460af9 Mon Sep 17 00:00:00 2001 From: Bryan Helmkamp Date: Fri, 22 May 2026 14:03:26 -0400 Subject: [PATCH] feat: expand Ask Fabro example prompts to detailed text Replace the four built-in Ask Fabro welcome prompts with more detailed prompt text covering errors, performance, run flow tracing, and workflow improvement recommendations. Card headings and descriptions are unchanged; only the verbatim prompt sent on click is expanded. Co-Authored-By: Claude Opus 4.7 (1M context) --- apps/fabro-web/app/components/chats/sidebar-welcome.tsx | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/apps/fabro-web/app/components/chats/sidebar-welcome.tsx b/apps/fabro-web/app/components/chats/sidebar-welcome.tsx index 1818f5340..c8e9f1fa7 100644 --- a/apps/fabro-web/app/components/chats/sidebar-welcome.tsx +++ b/apps/fabro-web/app/components/chats/sidebar-welcome.tsx @@ -9,25 +9,25 @@ const EXAMPLE_PROMPTS = [ heading: "Surface errors", description: "Find errors, warnings, and failed steps in this run.", prompt: - "Walk me through any errors, warnings, or failed steps in this run and what caused them.", + "Analyze this workflow run for any errors, exceptions, failed operations, or unexpected behavior. Include details about what went wrong, when it occurred, and potential causes or fixes.", }, { heading: "Analyze performance", description: "Spot the slowest stages and where time was spent.", prompt: - "Which stages were the slowest, and where did this run spend most of its time?", + "Analyze the performance of this workflow run, including tool execution times, API call latencies, timeouts, and any bottlenecks. Compare timing across similar operations and identify optimization opportunities.", }, { heading: "Review key decisions", description: "Recap the important choices the agent made.", prompt: - "What were the key decisions the agent made during this run, and why?", + "Trace the workflow run flow, including key decision points, branching logic, tool selection reasoning, and how the assistant responded to different inputs or contexts.", }, { heading: "Suggest improvements", description: "Ideas to make this workflow faster and more reliable.", prompt: - "Suggest improvements to make this workflow faster and more reliable.", + "Provide recommendations for improving this workflow, including better graph design, prompting strategies, more efficient tool usage, error handling improvements, and ways to optimize the overall user experience.", }, ];