amysguan 47b0c87ebe Add modes self and self_sliding_window for prompt caching (#9372) ...

* add self compaction method with proper caching (pass in tools, don't refresh sys prompt beforehand) + sliding fallback

* updated prompts for self compaction

* add tests for self, self_sliding_window modes and w/o refresh messages before compaction

* add cache logging to summarization

* better handling to prevent agent from continuing convo on self modes

* if mode changes via summarize endpoint, will use default prompt for the new mode

---------

Co-authored-by: Amy Guan <amy@letta.com>

2026-02-24 10:55:26 -08:00

..

adapters

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

configs

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

data

feat(core): add image support in tool returns [LET-7140] (#8985)

2026-01-29 12:43:53 -08:00

helpers

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

managers

fix: models need to be paginated (#9621)

2026-02-24 10:55:26 -08:00

mcp_tests

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

model_settings

feat: add glm 5 to core (#9436)

2026-02-24 10:52:07 -08:00

performance_tests

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

sdk

test: fix parallel tool call default value (#9572)

2026-02-24 10:55:12 -08:00

test_agent_files

fix(memory): standardize tool parameter names (#9552)

2026-02-24 10:55:24 -08:00

test_tool_sandbox

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_tool_schema_parsing_files

merge this (#4759)

2025-09-17 15:47:40 -07:00

__init__.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

clear_postgres_db.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

config.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

conftest.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

constants.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

integration_test_agent_tool_graph.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_async_tool_sandbox.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_batch_api_cron_jobs.py

fix: duplicate session commit in step logging (#7512)

2026-01-12 10:57:19 -08:00

integration_test_batch_sdk.py

feat: Fix test batch sdkpy [LET-4507] (#4917)

2025-10-07 17:50:46 -07:00

integration_test_builtin_tools.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

integration_test_cancellation.py

fix: use shared event + .athrow() to properly set stream_was_cancelle… (#9019)

2026-01-29 12:44:04 -08:00

integration_test_chat_completions.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_clickhouse_llm_traces.py

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

integration_test_client_side_tools.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

integration_test_composio.py

chore: Remove test composio (#4914)

2025-10-07 17:50:46 -07:00

integration_test_conversations_sdk.py

feat: add conversation deletion endpoint (soft delete) [LET-7286] (#9230)

2026-02-24 10:55:12 -08:00

integration_test_human_in_the_loop.py

fix: handle compact edge case in idempotency check (#9588)

2026-02-24 10:55:12 -08:00

integration_test_mcp.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_modal_sandbox_v2.py

feat: pass in LETTA_TOOL_ID, LETTA_AGENT_ID, LETTA_PROJECT_ID into tool env (#6439)

2025-12-15 12:02:18 -08:00

integration_test_modal.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_multi_agent.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

integration_test_multi_modal_tool_returns.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_override_model.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_send_message_v2.py

fix(core): add OpenAI 24h prompt cache retention for supported models (#9509)

2026-02-24 10:55:11 -08:00

integration_test_send_message.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

integration_test_sleeptime_agent.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_summarizer.py

Add modes self and self_sliding_window for prompt caching (#9372)

2026-02-24 10:55:26 -08:00

integration_test_system_prompt_prefix_caching.py

feat: Prefix cache optimization system prompt (#9381)

2026-02-24 10:52:07 -08:00

integration_test_token_counters.py

fix: duplicate session commit in step logging (#7512)

2026-01-12 10:57:19 -08:00

integration_test_tool_execution_sandbox.py

fix: duplicate session commit in step logging (#7512)

2026-01-12 10:57:19 -08:00

integration_test_turbopuffer.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

integration_test_typescript_tool_execution_sandbox.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

integration_test_usage_tracking.py

feat(gemini): add 3.1 pro preview support (#9553)

2026-02-24 10:55:11 -08:00

integration_test_voice_agent.py

feat: patch model listing to actually match handle [LET-5888] (#5754)

2025-11-13 15:35:34 -08:00

locust_test.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

manual_test_many_messages.py

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

manual_test_multi_agent_broadcast_large.py

chore: deprecate old agent messaging (#9120)

2026-02-24 10:52:06 -08:00

mcp_test.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

mock_mcp_server.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

pytest.ini

feat: modal tool execution - NO FEATURE FLAGS USES MODAL [LET-4357] (#5120)

2025-11-13 15:36:56 -08:00

test_agent_serialization_v2.py

fix(core): handle None message_ids in context window calculator (#9330)

2026-02-24 10:55:11 -08:00

test_agent_serialization.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_cli.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_client.py

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

test_context_window_calculator.py

fix(core): stabilize system prompt refresh and expand git-memory coverage (#9438)

2026-02-24 10:52:07 -08:00

test_crypto_utils.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_embedding_optional.py

fix: don't need embedding model for self hosted [LET-7009] (#8935)

2026-01-29 12:44:04 -08:00

test_embeddings.py

Revert "fix: truncate oversized text in embedding requests" (#9227)

2026-02-24 10:52:06 -08:00

test_exception_logging.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_file_processor.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_google_embeddings.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_google_schema_refs.py

fix: google gen ai format error fix (#9147)

2026-02-24 10:52:07 -08:00

test_internal_agents_count.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_letta_agent_batch.py

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

test_letta_request_schema.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_llm_clients.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_log_context_middleware.py

fix(core): preserve git-memory formatting and enforce lock conflicts (#9451)

2026-02-24 10:52:07 -08:00

test_log_context.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_long_running_agents.py

chore: bump wait_for_server from 30 to 60 seconds (#8435)

2026-01-12 10:57:49 -08:00

test_managers.py

chore: Remove test managers (#4926)

2025-10-07 17:50:46 -07:00

test_mcp_encryption.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_memory.py

Fix git-memory context preview parsing (#9414)

2026-02-24 10:52:07 -08:00

test_message_serialization.py

Revert "fix: filter orphaned approval_request messages to prevent Anthropic API errors" (#8721)

2026-01-19 15:54:39 -08:00

test_minimax_client.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_modal_sandbox_v2.py

feat: pass in LETTA_TOOL_ID, LETTA_AGENT_ID, LETTA_PROJECT_ID into tool env (#6439)

2025-12-15 12:02:18 -08:00

test_multi_agent.py

chore: deprecate old agent messaging (#9120)

2026-02-24 10:52:06 -08:00

test_openai_prompt_cache_request_fields.py

fix(core): add OpenAI 24h prompt cache retention for supported models (#9509)

2026-02-24 10:55:11 -08:00

test_optimistic_json_parser.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_otel_tracing.py

fix: prevent huge otel spans causing pods to be OOMKilled (#5871)

2025-11-13 15:36:20 -08:00

test_plugins.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_prompt_caching.py

feat(gemini): add 3.1 pro preview support (#9553)

2026-02-24 10:55:11 -08:00

test_provider_api.py

feat: allow suers to test their own key [LET-5312] (#5257)

2025-10-09 13:20:53 -07:00

test_provider_trace_agents.py

fix: add LLMCallType enum and ensure call_type is set on all provider traces (#9258)

2026-02-24 10:52:06 -08:00

test_provider_trace_backends.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_provider_trace_summarization.py

Add modes self and self_sliding_window for prompt caching (#9372)

2026-02-24 10:55:26 -08:00

test_provider_trace.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_providers.py

fix: add m2.5 (#9480)

2026-02-24 10:52:07 -08:00

test_redis_client.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_run_status_conversion.py

feat: add status to runs api [PRO-1529] (#5214)

2025-10-07 17:50:50 -07:00

test_schema_validator.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_sdk_client.py

test: remove sonnet 3-7 reference (#9618)

2026-02-24 10:55:26 -08:00

test_secret.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_server_providers.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_server.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_sonnet_nonnative_reasoning_buffering.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_sources.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_static_buffer_summarize.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_stream_buffer_readers.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_temporal_metrics_local.py

chore: enable F821, F401, W293 (#9503)

2026-02-24 10:55:08 -08:00

test_timezone_formatting.py

merge this (#4759)

2025-09-17 15:47:40 -07:00

test_tool_rule_solver.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_tool_schema_parsing.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

test_usage_parsing.py

fix: update Anthropic Haiku test model after 3.5 retirement (#9569)

2026-02-24 10:55:12 -08:00

test_utils.py

chore: add ty + pre-commit hook and repeal even more ruff rules (#9504)

2026-02-24 10:55:11 -08:00

utils.py

feat: cutover repo to 1.0 sdk client LET-6256 (#6361)

2025-11-24 19:11:18 -08:00