Skip to content

armgit / letta

Public Imported from GitHub
0

Platform for stateful agents: AI with advanced memory that can learn and self-improve over time.

fix: add tool name info to tool generation prompting (#3635)

cthomas3e9f7db175e9
diff --git a/letta/prompts/system/memgpt_generate_tool.txt b/letta/prompts/system/memgpt_generate_tool.txt
index 17a9353a8..730c60ad5 100644
--- a/letta/prompts/system/memgpt_generate_tool.txt
+++ b/letta/prompts/system/memgpt_generate_tool.txt
@@ -32,14 +32,15 @@ Monologues can reflect your thinking process, inner reflections, and personal gr
<tool_generation>
You are are expert python programmer that is tasked with generating python source code for tools that the user can use in their LLM invocations.
**Quick Rules for Generation**
-1. **Use a flat, one-line signature** with only native types:
+1. **Never rename** the provided function name, even if core functionality diverges. The tool name is a static property.
+2. **Use a flat, one-line signature** with only native types:
```python
def tool_name(param1: str, flag: bool) -> dict:
```
-2. **Docstring `Args:`** must list each parameter with a **single token** type (`str`, `bool`, `int`, `float`, `list`, `dict`).
-3. **Avoid** `Union[...]`, `List[...]`, multi-line signatures, or pipes in types.
-4. **Don't import NumPy** or define nested `def`/`class`/decorator blocks inside the function.
-5. **Simplify your `Returns:`**—no JSON-literals, no braces or `|` unions, no inline comments.
+3. **Docstring `Args:`** must list each parameter with a **single token** type (`str`, `bool`, `int`, `float`, `list`, `dict`).
+4. **Avoid** `Union[...]`, `List[...]`, multi-line signatures, or pipes in types.
+5. **Don't import NumPy** or define nested `def`/`class`/decorator blocks inside the function.
+6. **Simplify your `Returns:`**—no JSON-literals, no braces or `|` unions, no inline comments.
</tool_generation>
<tool_signature>