Fimeg/letta-code
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
1,086 Commits 1 Branch 0 Tags
f92279c2a58449a67fe16bf5a60d9e42cdf001a8
Commit Graph

578 Commits

Author SHA1 Message Date
Charles Packer
f92279c2a5 fix: compact plan and todo tool header args (#1031) 2026-02-18 23:31:59 -08:00
Charles Packer
366c1a39c7 fix: preserve footer reasoning tier across restart (#1030) 2026-02-18 23:13:02 -08:00
Charles Packer
47e81433ff feat: align TaskOutput UX with Bash output (#1029) 2026-02-18 22:53:41 -08:00
paulbettner
674e8c12c2 fix(tui): show elapsed time for running shell tools (#1026) 2026-02-18 20:01:47 -08:00
Charles Packer
dc25ce5573 fix: harden permissions matching and alias caching (#1027) 2026-02-18 19:49:53 -08:00
cpacker
6b7c59b0be fix(tui): populate reasoning_effort from updateArgs as fallback when API omits it 
The Letta API _to_model_settings() for Anthropic was not including
effort in the GET response (backend bug), so agentState.model_settings.effort
and llm_config.reasoning_effort both came back null after a /model switch.
deriveReasoningEffort then had nothing to work with.

Client-side fix: after updateAgentLLMConfig returns, merge reasoning_effort
from model.updateArgs (/model switch) and desired.effort (Tab cycle flush)
into the llmConfig state we set. This populates the fallback path in
deriveReasoningEffort reliably regardless of what the API echoes back.

The actual root cause is fixed in letta-cloud: _to_model_settings() now
includes effort=self.effort for Anthropic models.

👾 Generated with [Letta Code](https://letta.com)

Co-Authored-By: Letta <noreply@letta.com>
 2026-02-18 19:00:07 -08:00
Charles Packer
35291d9094 fix(tui): footer reasoning tier, toolset naming, and selector highlight (#1024) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-18 18:48:25 -08:00
paulbettner
b871ff793d fix(permissions): avoid stale YOLO mode during approvals (#1023) 2026-02-18 18:41:58 -08:00
Charles Packer
a0a5bbfc72 feat: inject toolset swap and slash-command context reminders (#1022) 2026-02-18 18:15:38 -08:00
jnjpng
f90244de82 fix: reset conversation routing after /agents new agent creation (#1020) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-18 17:44:07 -08:00
Shubham Naik
56ff8ad447 feat: listen mode (#997) 
Co-authored-by: cpacker <packercharles@gmail.com>
 2026-02-18 17:38:48 -08:00
paulbettner
3871fa810c fix(permissions): prevent yolo mode desync in approval handling (#1021) 2026-02-18 17:23:36 -08:00
Charles Packer
75b9615d47 feat(tui): cycle reasoning tiers with Tab key (#1019) 
Co-authored-by: paulbettner <paulbettner@playfulstudios.com>
 2026-02-18 16:52:49 -08:00
Charles Packer
96396f023d fix: execute tools against dispatch-time snapshot (#1018) 2026-02-18 16:49:54 -08:00
jnjpng
f3aa3291db fix: standardize slash command help and error messages (#999) 2026-02-18 16:06:54 -08:00
Devansh Jain
b322a46a43 refactor: Unify reminder management across interactive and headless modes (#1001) 
Co-authored-by: Letta <noreply@letta.com>
Co-authored-by: cpacker <packercharles@gmail.com>
 2026-02-18 15:44:59 -08:00
Charles Packer
5fac4bc106 feat: persist toolset mode and add auto option (#1015) 2026-02-18 13:43:40 -08:00
Charles Packer
bdc23932b5 fix: send Anthropic effort in model_settings and read it for display (#1016) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-18 12:57:52 -08:00
Charles Packer
4e1827ebc1 refactor: tighten codex toolset parity without freeform (#1014) 2026-02-18 12:55:09 -08:00
Charles Packer
b505ea6117 fix: improve model/reasoning display in footer and slash preview (#1008) 2026-02-17 22:23:47 -08:00
Charles Packer
245390adb0 feat: add reasoning settings step to /model (#1007) 2026-02-17 21:27:13 -08:00
paulbettner
9b438bdd91 fix(cli): let lowercase d type in /agents search (#1006) 2026-02-17 21:18:38 -08:00
Charles Packer
61139bfd4f fix(cli): keep model selector chrome visible during refresh (#1005) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-17 18:29:04 -08:00
Charles Packer
d71ce07c25 fix(cli): use white text on active tab highlight across all selectors (#1004) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-17 18:28:41 -08:00
Charles Packer
5d9f2b68ff fix(cli): reinject bootstrap reminders on conversation switches (#1000) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-17 16:45:28 -08:00
Kian Jones
cda9f3d71d fix: ADE links for default conversation (#998) 2026-02-17 15:43:57 -08:00
jnjpng
66f333a5ca fix: correctly display no arguments tip on commands that don't accept them instead of opening state (#996) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-17 15:14:17 -08:00
Charles Packer
3728d1ec0c refactor: extract shared turn queue runtime (#991) 2026-02-17 12:58:33 -08:00
Christina Tong
44d4cc87c1 feat: pass Edit code diff start line in tool return (#994) 2026-02-17 12:01:03 -08:00
Charles Packer
7cc8729e57 feat: align headless bidirectional queueing with TUI runtime (#990) 2026-02-17 11:38:29 -08:00
Charles Packer
6a2b2f6346 feat: add skill source controls and headless reminder settings (#989) 2026-02-16 23:01:03 -08:00
Charles Packer
e074d0bfb7 fix: gate skills reminders in headless sessions (#987) 2026-02-16 19:08:55 -08:00
Charles Packer
66ced52c81 feat: expose MEMORY_DIR and AGENT_ID in runtime context (#986) 2026-02-16 18:24:59 -08:00
Charles Packer
72c8a0ab23 fix: reconcile memfs/standard prompt sections safely (#985) 2026-02-16 17:29:52 -08:00
Charles Packer
2ac0b8b262 fix(cli): reset interrupt timer on eager cancel (#984) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-16 16:05:17 -08:00
Charles Packer
408bb8e81e fix: block slash command queueing while agent is running (#983) 2026-02-16 15:58:59 -08:00
Charles Packer
5435f44c45 fix: expand chatgpt oauth retry classification (#980) 2026-02-16 14:36:15 -08:00
jnjpng
f2f59e4591 fix(retry): broaden ChatGPT error pattern to catch all ChatGPT API errors (#978) 2026-02-16 13:40:16 -08:00
jnjpng
48a448dd59 fix(retry): add ChatGPT connection error patterns to retriable error detection (#977) 2026-02-16 13:29:24 -08:00
Charles Packer
8ded47e945 feat(cli): add /skills overlay to browse available skills (#961) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-14 11:45:33 -08:00
Charles Packer
e337330dbe fix(cli): resolve global LRU agent on directory switch instead of cre… (#951) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-13 23:02:38 -08:00
Kian Jones
81ac1670ab fix: resolve ghost assistant messages when Anthropic returns [text, thinking, text] (#956) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-13 17:44:46 -08:00
jnjpng
40c508fe8c feat: show specific retry messages for known LLM provider errors (#953) 2026-02-13 12:31:23 -08:00
Charles Packer
1f44612a01 refactor(cli): unify turn recovery policy between TUI and headless (#950) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-12 21:52:13 -08:00
Charles Packer
45bd0a6af9 fix: resolve CONFLICT error after interrupt during tool execution (#949) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-12 19:35:01 -08:00
Devansh Jain
415b1a403b feat: add --no-skills flag to disable bundled skills (#948) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-12 18:26:46 -08:00
Christina Tong
5711f2133e Christina/fix link (#945) 2026-02-12 17:49:24 -08:00
Charles Packer
0272a2e371 fix: default reflection trigger to compaction-event (#943) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-12 17:42:38 -08:00
Christina Tong
dc33b4cd28 fix: ade default convo link (#944) 2026-02-12 17:39:05 -08:00
Devansh Jain
36c21939a9 fix: add missing memfs init steps in headless mode + refactor (#941) 
Co-authored-by: Letta <noreply@letta.com>
 2026-02-12 17:33:51 -08:00