Fix using research tool names instead of slash command tool names
(exposed to user) in research mode conversation history construction.
Map agent input tools to relevant research tools. Previously
using agents with a limited set of tools in research mode reduces
tools available to agent in research mode.
Fix checks to skip tools if not configured.
The chat model friendly name field was introduced in a8c47a70f. But
we weren't setting the friendly name for ollama models, which get
automatically loaded on first run.
This broke setting chat model options, server admin settings and
creating new chat pages (at least) as they display the chat model's
friendly name.
This change ensures the friendly name for auto loaded chat models is
set to resolve these issues. We also add a null ref check to web app
model selector as an additional safeguard to prevent new chat page
crash due to missing friendly name going forward.
Resolves#1208
We'd reversed the formatting of openai messages to drop invalid
messages without affecting the other messages being appended . But we
need to reverse the final formatted list to return in the right order.
Previously
- message with invalid content were getting dropped in normal order
which would change the item index being iterated for gemini and
anthropic models
- messages with empty content weren't getting dropped for openai
compatible api models. While openai api is resilient to this, it's
better to drop these invalid messages as other openai compatible
APIs may not handle this.
We see messages with empty or no content when chat gets interrupted
due to disconnections, interrupt messages or explicit aborts by user.
This changes should now drop invalid messages and not mess formatting
of the other messages in a conversation. It should allow continuing
interrupted conversations with any ai model.
Inspired by my previous turnstyle ux explorations.
But basically user message becomes section title and khoj message
becomes section body with the timestamp being used a section title,
body divider.
Previous organic results enumerator only handled the scenario where
organic key wasn't present in online search results.
It did not handle the case where there were no organic online search
results.
- Methods calling send_message_to_model_wrapper_sync handn't been
update to handle the function returning new ResponseWithThought
- Store, load request.url to DB as, from string to avoid serialization
issues
For files not synced after the previous release, context uri is unset.
This results in failure to save chat messages that retrieve documents
as the uri field cannot be unset so pre save validation fails.
We'd use a db migration to handle this but this is a quick mitigation
for now.
For files not synced after the previous release, context uri is unset.
This results in failure to save chat messages that retrieve documents
as the uri field cannot be unset so pre save validation fails
Use url fragment schema for deep link URIs, borrowing from URL/PDF
schemas. E.g file:///path/to/file.txt#line=<line_no>&#page=<page_no>
Compute line number during (recursive) markdown entry chunking.
Test line number in URI maps to line number of chunk in actual md file.
This deeplink URI with line number is passed to llm as context to
better combine with line range based view file tool.
Grep tool already passed matching line number. This change passes
line number in URIs of markdown entries matched by the semantic search
tool.
Use url fragment schema for deep link URIs, borrowing from URL/PDF
schemas. E.g file:///path/to/file.txt#line=<line_no>&#page=<page_no>
Compute line number during (recursive) org-mode entry chunking.
Thoroughly test line number in URI maps to line number of chunk in
actual org mode file.
This deeplink URI with line number is passed to llm as context to
better combine with line range based view file tool.
Grep tool already passed matching line number. This change passes
line number in URIs of org entries matched by the semantic search tool
Only embedding models see, operate on compiled text.
LLMs should see raw entry to improve combining it with other document
traversal tools for better regex and line matching.
Users see raw entry for better matching with their actual notes.
Reduce structural changes to raw entry allows better deep-linking and
re-annotation. Currently done via line number in new uri field.
Only add properties drawer to raw entry if entry has properties
Previously line and source properties were inserted into raw entries.
This isn't done anymore. Line, source are deprecated for use in khoj.el.
- Cache last anthropic message. Given research mode now uses function
calling paradigm and not the old research mode structure.
- Cache tool definitions passed to anthropic models
- Stop dropping first message if by assistant as seems like Anthropic
API doesn't complain about it any more.
- Drop tool result when tool call is truncated as invalid state
- Do not truncate tool use message content, just drop the whole tool
use message.
AI model APIs need tool use assistant message content in specific
form (e.g with thinking etc.). So dropping content items breaks
expected tool use message content format.
Handle tool use scenarios where iteration query isn't set for retry
- Deepcopy messages before formatting message for Anthropic to allow
idempotency so retry on failure behaves as expected
- Handle failed calls to pick next tools to pass failure warning and
continue next research iteration. Previously if API call to pick
next failed, the research run would crash
- Add null response check for when Gemini models fail to respond
Previously if anthropic models were using tools, the models text
response accompanying the tool use wouldn't be shown as they were
overwritten in aggregated response with the tool call json.
This changes appends the text response to the thoughts portion on tool
use to still show model's thinking. Thinking and text response are
delineated by italics vs normal text for such cases.
This should avoid the need to reformat the Khoj standardized tool call
for cache hits and satisfying ai model api requirements.
Previously multi-turn tool use calls to anthropic reasoning models
would fail as needed their thoughts to be passed back. Other AI model
providers can have other requirements.
Passing back the raw response as is should satisfy the default case.
Tracking raw response should make it easy to apply any formatting
required before sending previous response back, if any ai model
provider requires that.
Details
---
- Raw response content is passed back in ResponseWithThoughts.
- Research iteration stores this and puts it into model response
ChatMessageModel when constructing iteration history when it is
present.
Fallback to using parsed tool call when raw response isn't present.
- No need to format tool call messages for anthropic models as we're
passing the raw response as is.
Researcher is expanding into accomplish task behavior, especially with
tool use from the previous collect information to answer user query
behavior.
Update the researcher's system prompt to reflect the new objective better.
Encourage model to not stop working on task until achieve objective
Earlier khoj could technically only answer questions existential
questions, i.e question that would terminate once any relevant note to
answer that question was found.
This change enables khoj to answer universal questions, i.e questions
that require searching through all notes or finding all instances.
It enables more thorough retrieval from user's knowledge base by
combining semantic search, regex search, view and list files tools.
For more development details including motivation, see live coding
session 1.1 at https://www.youtube.com/live/-2s_qi4hd2k
Allow getting a map of user's knowledge base under specified path.
This enables more thorough retrieval from user's knowledge base by
combining search, view and list files tools.
