Skip to content

Agent Zero AI framework

Reinforce response tool usage in prompts

linkliti626a44696c0e
diff --git a/prompts/agent.system.main.communication.md b/prompts/agent.system.main.communication.md
index b200b553a..c43795d50 100644
--- a/prompts/agent.system.main.communication.md
+++ b/prompts/agent.system.main.communication.md
@@ -1,5 +1,6 @@
## Communication
+- Always reply through a tool call; plain-text output is invalid and never reaches the user
- Output must be valid JSON with double quotes for all keys and string values
- No JSON in markdown fences
- Do not invent unavailable tool names and args
diff --git a/prompts/agent.system.response_tool_tips.md b/prompts/agent.system.response_tool_tips.md
index 1bee85f00..13f77327d 100644
--- a/prompts/agent.system.response_tool_tips.md
+++ b/prompts/agent.system.response_tool_tips.md
@@ -1 +1,2 @@
+always deliver the final user-facing answer through this `response` tool's `text` arg; plain-text output without a tool call is invalid and never reaches the user
for long existing text, use `§§include(path)` instead of rewriting