Make researcher handle ambiguous requests better by working with
reasonable assumptions (clearly told to user in response) instead of
burdering user with clarification requests.
Fix portions of the researcher prompt that had gone stale since moving
to tool use and making researcher more task (vs q&a) oriented
Previously the researcher was passing the whole code to execute in its
queries to the tool AI instead of asking it to write the code and
limiting its query to a natural language request (with required data).
The division of responsibility should help researcher just worry about
constructing a request with all the required details instead of also
worrying about writing correct code.
Their tool call response may not strictly follow expected response
format. Let researcher handle incorrect arguments to code tool (i.e
triggers type error)
What
- Get reasoning of openai reasoning models from responses api for sho
- Improves cache hits and reasoning reuse for iterative agents like
research mode.
This should improve speed, quality, cost and transparency of using
openai reasoning models.
More cache hits and better reasoning as reasoning blocks are included
while model is researching (reasoning intersperse with tool calls)
when using the responses api.
Previously line start, end anchors would just work if the whole file
started or ended with the regex pattern rather than matching by line.
Fix it to work like a standard grep tool and match by line start, end.
Reduce usage of boolean operators like "hello OR bye OR see you" which
doesn't work and reduces search quality. They're trying to stuff the
search query with multiple different queries.
## Overview
Speed up app install and development using a faster, modern development
toolchain
## Details
### Major
- Use [uv](https://docs.astral.sh/uv/) for faster server install (vs
pip)
- Use [bun](https://bun.sh/) for faster web app install (vs yarn)
- Use [ruff](https://docs.astral.sh/ruff/) for faster formatting of
server code (vs black, isort)
- Fix devcontainer builds. See if uv and bun can speed up server and
client installs
### Minor
- Format web app with prettier and server with ruff. This is most of the
file changes in this PR.
- Simplify copying web app built files in pypi workflow to make it less
flaky.
- CI runners don't have GPUs
- Pytorch related Nvidia cuda packages are not required for testing,
evals or pre-commit checks.
- Avoiding these massive downloads should speed up workflow run.
### Overview
Make server leaner to increase development speed.
Remove old indexing code and the native offline chat which was hard to
maintain.
- The native offline chat module was written when the local ai model api
ecosystem wasn't mature. Now it is. Reuse that.
- Offline chat requires GPU for usable speeds. Decoupling offline chat
from Khoj server is the recommended way to go for practical inference
speeds (e.g Ollama on machine, Khoj in docker etc.)
### Details
- Drop old code to index files on server filesystem. Clean cli, init
paths.
- Drop native offline chat support with llama-cpp-python.
Use established local ai APIs like Llama.cpp Server, Ollama, vLLM etc.
- Drop old pre 1.0 khoj config migration scripts
- Update test setup to index test data after old indexing code removed.
- Delete tests testing deprecated server side indexing flows
- Delete `Local(Plaintext|Org|Markdown|Pdf)Config' methods, files and
references in tests
- Index test data via new helper method, `get_index_files'
- It is modelled after the old `get_org_files' variants in main app
- It passes the test data in required format to `configure_content'
Allows maintaining the more realistic tests from before while
using new indexing mechanism (rather than the deprecated server
side indexing mechanism
This stale code was originally used to index files on server file
system directly by server. We currently push files to sync via API.
Server side syncing of remote content like Github and Notion is still
supported. But old, unused code for server side sync of files on
server fs is being cleaned out.
New --log-file cli args allows specifying where khoj server should
store logs on fs. This replaces the --config-file cli arg that was
only being used as a proxy for deciding where to store the log file.
- TODO
- Tests are broken. They were relying on the server side content
syncing for test setup
It is recommended to chat with open-source models by running an
open-source server like Ollama, Llama.cpp on your GPU powered machine
or use a commercial provider of open-source models like DeepInfra or
OpenRouter.
These chat model serving options provide a mature Openai compatible
API that already works with Khoj.
Directly using offline chat models only worked reasonably with pip
install on a machine with GPU. Docker setup of khoj had trouble with
accessing GPU. And without GPU access offline chat is too slow.
Deprecating support for an offline chat provider directly from within
Khoj will reduce code complexity and increase developement velocity.
Offline models are subsumed to use existing Openai ai model provider.
Clarify that the tool AI will perform a maximum of X sub-queries for
each query passed to it by the manager AI.
Avoids the manager AI from trying to directly pass a list of queries
to the search tool AI. It should just pass just a single query.
Send larger thought chunks to improve streaming efficiency and
reduce rendering load on web client.
This rendering load was most evident when using high throughput
models or low compute clients.
The server side message buffering should result in fewer re-renders,
faster streaming and lower compute load on client.
Related commit to buffer message content in fc99f8b37
- Ask both manager and code gen AI to not run or write
unsafe code for some safety improvement (over code exec in sandbox).
- Disallow custom agent prompts instructing unsafe code gen
## PR Summary
This PR resolves the deprecation warnings of the Pydantic library, which
you can find in the [CI
logs](https://github.com/khoj-ai/khoj/actions/runs/16528997676/job/46749452047#step:9:142):
```python
PydanticDeprecatedSince20: The `copy` method is deprecated; use `model_copy` instead. See the docstring of `BaseModel.copy` for details about how to handle `include` and `exclude`. Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.11/migration/
```
Save to conversation in normal flow should only be done if
interrupt wasn't triggered.
Saving conversations on interrupt is handled completely by the
disconnect monitor since the improvements to interrupt.
This abort is handled correctly for steps before final response. But
not if interrupt occurs while final response is being sent. This
changes checks for cancellation after final response send attempt and
avoids duplicate chat turn save.
- Extract llm thoughts from more openai compatible ai api providers
like llama.cpp server vllm and litellm.
- Try structured thought extraction by default
- Try in-stream thought extraction for specific model families like
qwen and deepseek.
- Show thoughts with tool use. For intermediate steps like research
mode from openai compatible models
Some consensus on thought in model response is being reached with
using deepseek style thoughts in structured response (via
"reasoning_content" field) or qwen style thoughts in main
response (i.e <think></think> tags).
Default to try deepseek style structured thought extraction. So the
previous default stream processor isn't required.
A previous regression resulted in the start llm response event being
sent with every (non-thought) message chunk. It should only be sent
once after thoughts and before first normal message chunk is streamed.
Regression probably introduced with changes to stream thoughts.
This should fix the chat streaming latency logs.
