From 54b39edaeae82c9575d1abeda5033d2a8f338fe1 Mon Sep 17 00:00:00 2001 From: akhisud3195 Date: Tue, 16 Sep 2025 16:55:19 +0400 Subject: [PATCH 1/3] Revert auth related changes to start.sh --- start.sh | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/start.sh b/start.sh index c98c240e..2ee2f895 100755 --- a/start.sh +++ b/start.sh @@ -22,17 +22,6 @@ export USE_KLAVIS_TOOLS=true # export USE_KLAVIS_TOOLS=true # fi -# default to disabling auth if not explicitly enabled -export USE_AUTH="${USE_AUTH:-false}" - -# provide dummy auth0 env vars if missing (to silence build-time warnings) -# Note: app/lib/auth0.ts expects AUTH0_ISSUER_BASE_URL and AUTH0_BASE_URL -export AUTH0_ISSUER_BASE_URL="${AUTH0_ISSUER_BASE_URL:-${AUTH0_DOMAIN:-test}}" -export AUTH0_CLIENT_ID="${AUTH0_CLIENT_ID:-test}" -export AUTH0_BASE_URL="${AUTH0_BASE_URL:-${APP_BASE_URL:-test}}" -export AUTH0_SECRET="${AUTH0_SECRET:-test}" -export AUTH0_CLIENT_SECRET="${AUTH0_CLIENT_SECRET:-test}" - # Start with the base command and profile flags CMD="docker compose" CMD="$CMD --profile setup_qdrant" From c2f3d87074779d4b731fe919c48250714a3b9ad4 Mon Sep 17 00:00:00 2001 From: tusharmagar Date: Tue, 16 Sep 2025 19:06:01 +0530 Subject: [PATCH 2/3] removed generate image tool from customer support prebuilt template --- .../lib/prebuilt-cards/customer-support.json | 22 ------------------- 1 file changed, 22 deletions(-) diff --git a/apps/rowboat/app/lib/prebuilt-cards/customer-support.json b/apps/rowboat/app/lib/prebuilt-cards/customer-support.json index a24d3b88..05668dfa 100644 --- a/apps/rowboat/app/lib/prebuilt-cards/customer-support.json +++ b/apps/rowboat/app/lib/prebuilt-cards/customer-support.json @@ -61,28 +61,6 @@ } ], "tools": [ - { - "name": "Generate Image", - "description": "Generate an image using Google Gemini given a text prompt. Returns base64-encoded image data and any text parts.", - "parameters": { - "type": "object", - "properties": { - "prompt": { - "type": "string", - "description": "Text prompt describing the image to generate" - }, - "modelName": { - "type": "string", - "description": "Optional Gemini model override" - } - }, - "required": [ - "prompt" - ], - "additionalProperties": true - }, - "isGeminiImage": true - }, { "name": "Mock Delivery Status", "description": "A mock tool to simulate checking delivery status.", From 7c261c6886974baff6f6e9e2d6ca112a3207485f Mon Sep 17 00:00:00 2001 From: arkml Date: Tue, 16 Sep 2025 20:01:34 +0530 Subject: [PATCH 3/3] make copilot aware of the internal image generation tool (#266) --- .../lib/copilot/copilot_multi_agent_build.ts | 6 ++- .../lib/copilot/example_multi_agent_1.ts | 51 +++++++++++++++++++ 2 files changed, 55 insertions(+), 2 deletions(-) diff --git a/apps/rowboat/src/application/lib/copilot/copilot_multi_agent_build.ts b/apps/rowboat/src/application/lib/copilot/copilot_multi_agent_build.ts index 8660fbcb..b5a7b215 100644 --- a/apps/rowboat/src/application/lib/copilot/copilot_multi_agent_build.ts +++ b/apps/rowboat/src/application/lib/copilot/copilot_multi_agent_build.ts @@ -189,6 +189,8 @@ Note: the rag_search tool searches across all data sources - it cannot call a sp 2. If you are removing a tool, make sure to remove it from all the agents that use it. 3. If you are adding a tool, make sure to add it to all the agents that need it. +Note: The agents have access to a tool called 'Generate Image'. This won't show up in the workflow like other tools. This tool can be used to generate images. If you want to add this tool to the agent, you can add it directly to the agent instructions like [@tool:Generate Image](#mention). + @@ -229,7 +231,7 @@ When a hub agent needs to execute a pipeline, it should: - + The user will provide the current config of the multi-agent system and ask you to make changes to it. Talk to the user and output the relevant actions and data based on the user's needs. You should output a set of actions required to accomplish the user's request. @@ -257,5 +259,5 @@ Below are details you should use when a user asks questions on how to use the pr {USING_ROWBOAT_DOCS} - + `; \ No newline at end of file diff --git a/apps/rowboat/src/application/lib/copilot/example_multi_agent_1.ts b/apps/rowboat/src/application/lib/copilot/example_multi_agent_1.ts index 9e7ba948..c94e5d97 100644 --- a/apps/rowboat/src/application/lib/copilot/example_multi_agent_1.ts +++ b/apps/rowboat/src/application/lib/copilot/example_multi_agent_1.ts @@ -1274,5 +1274,56 @@ Once you review and apply the changes, you can try out a basic chat first. I can --- +### Example 9: Image Generator Agent + +**User Request:** +Build an agent to generate images for me. + +**Copilot Response:** + +I'll create an agent that generates images for you. This agent will interact directly with you, ask for your image description, and use the built-in generate_image tool to create images based on your requests. + +Here’s what I’ll do: + +Create a user-facing agent called "Image Generator" that: +- Asks you for a description of the image you want. +- Uses the generate_image tool to generate the image. +- Returns the generated image to you. +Set clear instructions and examples for the agent. +(No need to add a new tool, as generate_image is built-in.) + +--- + +### 1. Create Agent + +I'll create an agent that generates images for you. This agent will interact directly with you, ask for your image description, and use the built-in generate_image tool to create images based on your requests. + +\`\`\`copilot_change +// action: create_new +// config_type: agent +// name: Image Generator + { + "change_description": "Created the image generator agent.", + "config_changes": { + "name": "Image Generator", + "type": "conversation", + "description": "Generates images for users based on their descriptions.", + "disabled": false, + "instructions": "## 🧑‍💼 Role:\nYou are an internal agent that generates images for users based on their descriptions.\n\n---\n## ⚙️ Steps to Follow:\n1. Receive the image description from the parent agent.\n2. Determine if the description is about an image.\n - If yes: Use the [@tool:Generate Image](#mention) tool to generate an image based on the user's description.\n - If not: Return control to the parent agent with a message such as \"This description is not about an image. Returning to parent agent.\"\n3. Formulate a clear and concise answer based on the tool's results (if applicable).\n4. If question is out of scope call [@agent:Image Generator](#mention) \n---\n## 🎯 Scope:\n✅ In Scope:\n- Generating images based on user descriptions.\n- Returning control to parent if the description is out of scope.\n\n❌ Out of Scope:\n- Handling any other questions or tasks.\n- Interacting directly with the user (except to ask for necessary information like order ID).\n\n---\n## 📋 Guidelines:\n✔️ Dos:\n- Use the Generate Image tool to generate an image based on the user's description.\n- If the description is not about an image, return control to the parent agent with a clear message.\n- Ask for order details if needed.\n\n🚫 Don'ts:\n- Do not answer questions outside of image generation.\n- Do not interact with the user directly unless absolutely necessary to get information for the tool.\n- Do not ignore out-of-scope questions; always return to parent.\n", + "model": "gpt-4.1", + "locked": false, + "toggleAble": true, + "ragReturnType": "chunks", + "ragK": 3, + "outputVisibility": "user_facing", + "controlType": "retain", + "maxCallsPerParentAgent": 3 + } +} +\`\`\` + +--- + +Once you review and apply the changes, you can try chatting with the "Image Generator" agent to generate images from your descriptions. Would you like to set this new agent as your start agent, or keep your current one? `; \ No newline at end of file