Sarah Wooders
08ef32ace2
feature/anthropic-system-ephemeral-caching ( #2684 )
2025-06-18 10:10:59 -07:00
Astrotalk
1531829cc3
- Add cache control to system messages
...
- Cache static system content to reduce token costs on repeated requests
- Maintain existing conversation flow while optimizing API costs
- Support both string and list format system messages
This change leverages Anthropic's ephemeral caching to reduce costs for
frequently used system prompts and personas without affecting functionality.
2025-06-17 18:25:07 +05:30
Sarah Wooders
12146579c4
chore: bugfixes ( #2682 )
2025-06-15 14:59:23 -07:00
Sarah Wooders
32911892f2
fix: fix poetry.lock
2025-06-15 11:20:50 -07:00
Sarah Wooders
b2a4f7e173
Merge branch 'main' into 0.8.4-release-2
2025-06-15 11:19:51 -07:00
jnjpng
729acaedb2
feat: add token column to mcp_servers and pipe through to sse server config ( #2775 )
...
Co-authored-by: Jin Peng <jinjpeng@Jins-MacBook-Pro.local >
2025-06-14 18:48:10 -07:00
Matthew Zhou
c71a6d1917
feat: Fall back to Optimistic JSON parser if Pydantic one fails ( #2820 )
2025-06-14 14:29:59 -07:00
Matthew Zhou
093f65e891
feat: Change reset messages to always preserve the original system message ( #2813 )
2025-06-13 18:58:27 -07:00
Sarah Wooders
78e81391ba
fix: avoid calling model_dump on stop reason messages twice ( #2811 )
2025-06-13 18:25:35 -07:00
Matthew Zhou
96a6e21dd6
fix: Relax constraints on pip requirements ( #2809 )
2025-06-13 17:21:04 -07:00
Matthew Zhou
17fc3420ed
fix: Make voice beta endpoint VAPI compatible ( #2808 )
2025-06-13 17:20:25 -07:00
Kevin Lin
8148dab8a8
feat: tune file tools prompting ( #2652 )
2025-06-13 16:24:59 -07:00
cthomas
97986b0f03
feat: send stop reason in letta APIs ( #2789 )
2025-06-13 16:04:48 -07:00
Kevin Lin
a8895ad017
feat: add reasoning models to integration_test_send_message ( #2710 )
2025-06-13 14:54:37 -07:00
Andy Li
562a404e27
fix: metrics imports ( #2801 )
2025-06-13 14:39:24 -07:00
Andy Li
bf7bd3b905
fix: metric tracking ( #2785 )
2025-06-13 13:53:10 -07:00
Matthew Zhou
b6affd1a16
feat: Add optional pip requirements to tool object ( #2793 )
2025-06-13 13:20:36 -07:00
Andy Li
0e45a90609
chore: workflow consolidatation and fern caching for tests on pr ( #2787 )
2025-06-13 12:33:53 -07:00
Kian Jones
aeadd1aebe
fix(ci): autosummarize test fails (pydantic version dependant?) ( #2792 )
2025-06-13 12:16:48 -07:00
Andy Li
336896dc5c
feat: redis user caching ( #2774 )
2025-06-12 17:32:07 -07:00
Shangyin Tan
22b640a5dd
fix: adjust memory summary's recall message count ( #2778 )
2025-06-12 17:11:36 -07:00
cthomas
8f13c078ef
feat: add stop reason object ( #2783 )
2025-06-12 16:54:29 -07:00
cthomas
ccb203b264
feat: consolidate logic for finish tokens ( #2779 )
2025-06-12 15:24:06 -07:00
Kevin Lin
916fe6c1da
fix: use max_completion_tokens in chat completion request ( #2764 )
2025-06-12 10:58:23 -07:00
Sarah Wooders
6be55bd982
chore: 0.8.4 release ( #2680 )
2025-06-12 09:33:25 -07:00
Sarah Wooders
af70e7c9e3
chore: bump version 0.8.4
2025-06-11 23:20:42 -07:00
Sarah Wooders
b1949c7d17
Merge branch 'main' into 0.8.4-release
2025-06-11 23:18:44 -07:00
Charles Packer
5498061e58
docs: add vibecoding prompts to docs site ( #2579 )
...
Co-authored-by: Sarah Wooders <sarahwooders@gmail.com >
2025-06-11 22:10:17 -07:00
Matthew Zhou
28fdaaff74
fix: Fix claude sonnet 3_7 tests ( #2767 )
2025-06-11 17:53:03 -07:00
Kevin Lin
78dad0db7a
fix: patch reasoning models ( #2703 )
...
Co-authored-by: Charles Packer <packercharles@gmail.com >
2025-06-11 17:20:04 -07:00
Matthew Zhou
d0c2ef89ea
feat: Add errors when tool call violates tool rules ( #2766 )
2025-06-11 17:12:39 -07:00
cthomas
03f4867cbe
fix: composio tools with array args ( #2762 )
2025-06-11 16:24:37 -07:00
cthomas
cf357b931d
feat: bump letta client version ( #2759 )
2025-06-11 13:39:02 -07:00
cthomas
aedc95056a
test: add new agent fixture to send message test ( #2758 )
2025-06-11 11:48:34 -07:00
cthomas
654f6b5053
feat: make tool calls required for model proxy ( #2756 )
2025-06-11 11:36:45 -07:00
cthomas
f3bee4658c
fix: not enough values to unpack error ( #2753 )
2025-06-11 11:05:46 -07:00
Matthew Zhou
d616a99916
fix: Get rid of lru cache ( #2748 )
2025-06-11 10:44:36 -07:00
cthomas
1bf54e1d6c
fix: test managers failure ( #2754 )
2025-06-11 10:41:03 -07:00
Matthew Zhou
ba3d59bba5
feat: Add prompting to guide tool rule usage ( #2742 )
2025-06-10 16:21:27 -07:00
Matthew Zhou
8ced9e57ba
feat: Fix test managers and patch default org id ( #2746 )
2025-06-10 16:20:36 -07:00
cthomas
883050e761
feat: set request heartbeat for max steps ( #2739 )
2025-06-10 15:26:07 -07:00
cthomas
484a6f1d37
fix: parallel tool calling OpenAI ( #2738 )
2025-06-10 14:27:01 -07:00
cthomas
6a23bc743d
feat: support multi content part input ( #2717 )
2025-06-10 13:36:17 -07:00
Matthew Zhou
530f9488b6
fix: Turn parallel tool calling off for OpenAI ( #2737 )
2025-06-10 13:27:00 -07:00
Matthew Zhou
96b8f155d8
fix: Turn off parallel tool calling for Claude ( #2736 )
2025-06-10 13:04:20 -07:00
cthomas
142a4f0ed8
fix: async loop eligibility check ( #2727 )
2025-06-10 12:56:01 -07:00
Matthew Zhou
a5ee49e579
feat: Remove debug artifacts ( #2734 )
2025-06-10 12:26:45 -07:00
Matthew Zhou
c4d7abb0ff
feat: Cache actor calls ( #2732 )
2025-06-10 12:15:13 -07:00
Andy Li
9b91c94d48
feat: otel endpoint tracking for messages ( #2715 )
2025-06-10 11:02:36 -07:00
Andy Li
617d208bcc
fix: list ollama embeddings models ( #2711 )
2025-06-10 10:44:26 -07:00
