feat: Add intelligent conversation history for pronoun/reference resolution

[iamrajiv]

Fixes: #58

How It Works

                        BEFORE (No History)
+-------------------------------------------------------------------------+
|  User: "What did Adrian say about spiders?"                             |
|  Answer: "Adrian mentioned spiders have eight legs..."                  |
|                                                                         |
|  User: "What else did he say about it?"                                 |
|         |                                                               |
|         v                                                               |
|  LLM: "he"=? "it"=? ---> [X] Cannot resolve references                  |
+-------------------------------------------------------------------------+

                        AFTER (With History)
+-------------------------------------------------------------------------+
|                                                                         |
|  +---------------+    +------------------+    +-----------------------+ |
|  | User Query    |--->| History Context  |--->| LLM Query Translation | |
|  |               |    | (last 3-5 Q&A)   |    |                       | |
|  | "What else    |    |                  |    | Prompt includes:      | |
|  |  did he say   |    | Q1: What did     |    | - User query          | |
|  |  about it?"   |    |     Adrian say   |    | - Recent Q&A pairs    | |
|  |               |    |     about        |    |                       | |
|  +---------------+    |     spiders?     |    | LLM resolves:         | |
|                       | A1: Adrian       |    | "he" -> Adrian        | |
|                       |     mentioned    |    | "it" -> spiders       | |
|                       |     spiders...   |    |                       | |
|                       +------------------+    +-----------+-----------+ |
|                                                           |             |
|                                                           v             |
|                                               +-----------------------+ |
|                                               | Structured Search     | |
|                                               | Query (JSON)          | |
|                                               |                       | |
|                                               | entities: ["Adrian"]  | |
|                                               | topics: ["spiders"]   | |
|                                               +-----------------------+ |
|                                                           |             |
|                                                           v             |
|                                                    Correct Answer       |
+-------------------------------------------------------------------------+

Data Flow

+-------------+     +--------------------+     +---------------------+
|             |     |                    |     |                     |
| User Query  |---->| process_query()    |---->| search_conversation |
|             |     |                    |     | _with_language()    |
+-------------+     | 1. Get history     |     |                     |
                    |    prompt section  |     | model_instructions  |
                    | 2. Inject into     |     | contains history    |
                    |    lang_options    |     | context             |
                    +--------------------+     +----------+----------+
                                                          |
                    +--------------------+                v
                    |                    |     +---------------------+
                    | ConversationHistory|<----| search_query_from   |
                    |                    |     | _language()         |
                    | Stores Q&A after   |     |                     |
                    | answer generated   |     | LLM sees history    |
                    +--------------------+     | in prompt_preamble  |
                                               +---------------------+

[iamrajiv]

hi @gvanrossum this is my pr related to #58 i have tested this locally and below are the results for make format check test could you please review the pr once?

➜  typeagent-py git:(history) ✗ make format check test
.venv/bin/black -tpy312 -tpy313 -tpy314 typeagent test tools gmail demo
All done! ✨ 🍰 ✨
112 files left unchanged.
.venv/bin/pyright --pythonpath .venv/bin/python typeagent test tools gmail
WARNING: there is a new pyright version available (v1.1.406 -> v1.1.407).
Please install the new version or set PYRIGHT_PYTHON_FORCE_VERSION to `latest`

0 errors, 0 warnings, 0 informations
.venv/bin/pytest
============================= test session starts ==============================
platform darwin -- Python 3.13.9, pytest-8.4.2, pluggy-1.6.0
rootdir: /Users/rajiv.singh/Desktop/projects/tt/typeagent-py
configfile: pyproject.toml
testpaths: test
plugins: mock-3.15.1, asyncio-1.2.0, anyio-4.11.0, logfire-4.14.2
asyncio: mode=Mode.STRICT, debug=False, asyncio_default_fixture_loop_scope=function, asyncio_default_test_loop_scope=function
collected 358 items

test/test_add_messages_with_indexing.py ...                              [  0%]
test/test_auth.py .......                                                [  2%]
test/test_collections.py ...........................                     [ 10%]
test/test_conversation_metadata.py ....................                  [ 15%]
test/test_demo.py s                                                      [ 16%]
test/test_embedding_consistency.py ...                                   [ 17%]
test/test_embeddings.py ...........                                      [ 20%]
test/test_factory.py ..s                                                 [ 20%]
test/test_incremental_index.py ..                                        [ 21%]
test/test_interfaces.py ......................                           [ 27%]
test/test_knowledge.py s...                                              [ 28%]
test/test_kplib.py ......                                                [ 30%]
test/test_mcp_server.py s.                                               [ 31%]
test/test_message_text_index_population.py .                             [ 31%]
test/test_message_text_index_serialization.py ..                         [ 31%]
test/test_messageindex.py ....s...                                       [ 34%]
test/test_online.py s                                                    [ 34%]
test/test_podcast_incremental.py ..                                      [ 34%]
test/test_podcasts.py s                                                  [ 35%]
test/test_property_index_population.py s                                 [ 35%]
test/test_propindex.py ............                                      [ 38%]
test/test_query.py ..................................                    [ 48%]
test/test_query_method.py ss                                             [ 48%]
test/test_related_terms_fast.py .                                        [ 49%]
test/test_related_terms_index_population.py s                            [ 49%]
test/test_reltermsindex.py ........s....                                 [ 53%]
test/test_searchlib.py ................................................. [ 66%]
......                                                                   [ 68%]
test/test_secindex.py ...                                                [ 69%]
test/test_secindex_storage_integration.py .                              [ 69%]
test/test_semrefindex.py ......................                          [ 75%]
test/test_serialization.py ........                                      [ 77%]
test/test_sqlite_indexes.py .....................                        [ 83%]
test/test_sqlitestore.py .....                                           [ 85%]
test/test_storage_providers_unified.py ..............................    [ 93%]
test/test_timestampindex.py .                                            [ 93%]
test/test_transcripts.py ......s                                         [ 95%]
test/test_utils.py ..s.                                                  [ 96%]
test/test_vectorbase.py ...........                                      [100%]

======================= 344 passed, 14 skipped in 3.06s ========================

[gvanrossum]

Thank you @iamrajiv. I have restarted the tests -- they only run when started by someone with repo write access.

I'm traveling all of December, and I'm not sure when I can review this thoroughly.

[iamrajiv]

no worries @gvanrossum thanks alot