Charles Packer
|
f61980e5f6
|
fix: align SDK permission-mode args with CLI behavior (#33)
|
2026-02-10 16:12:16 -08:00 |
|
Sarah Wooders
|
ee64be00ed
|
Add custom tools support
Add custom tools support
- Add AgentTool, AgentToolResult, AnyAgentTool types
- Add tools option to SessionOptions
- Add tool-helpers.ts with jsonResult, readStringParam, etc.
- Implement tool registration via register_external_tools control request
- Handle execute_external_tool requests in stream()
🐾 Generated with [Letta Code](https://letta.com)
|
2026-02-10 11:30:18 -08:00 |
|
Cameron
|
b0b82a2a3d
|
feat: expose stop_reason on SDKResultMessage (#30)
|
2026-02-10 10:13:56 -08:00 |
|
Christina Tong
|
c23bf8ee0d
|
chore: add embedding model to createAgent (#26)
|
2026-02-06 16:18:48 -08:00 |
|
Christina Tong
|
5f37e513c8
|
fix: memory blocks config on createAgent (#21)
|
2026-02-03 16:44:05 -08:00 |
|
Christina Tong
|
4cb8af7be8
|
feat: refactor options types for createAgent and createSession (#20)
|
2026-02-03 15:27:32 -08:00 |
|
Cameron
|
fa4a6340e7
|
feat: add image support to send() (#14)
Co-authored-by: Letta <noreply@letta.com>
|
2026-02-02 17:59:40 -08:00 |
|
Christina Tong
|
0e657e5109
|
feat: add createAgent and refactor (#13)
|
2026-01-30 17:03:47 -08:00 |
|
Christina Tong
|
c4c5106797
|
fix: updatedInput support and clean up types (#6)
|
2026-01-28 12:28:48 -08:00 |
|
cpacker
|
ee7cc92724
|
Initial release of Letta Code SDK
Programmatic control of Letta Code CLI with persistent agent memory.
Features:
- createSession() / resumeSession() / prompt() API
- resumeConversation() for multi-threaded conversations
- Multi-turn conversations with memory
- Tool execution (Bash, Read, Edit, etc.)
- System prompt and memory configuration
- Permission callbacks (canUseTool)
- Message streaming with typed events
👾 Generated with [Letta Code](https://letta.com)
|
2026-01-27 00:30:09 -08:00 |
|