github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 05:05:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #11830] Tool calling for non python devs #7851

New Issue
Closed
opened 2026-04-12 20:01:00 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-12 20:01:00 -05:00
Owner
Copy Link

Originally created by @TehWardy on GitHub (Aug 9, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/11830

I'm loving that Ollama spins me up a HTTP endpoint that I can use to interact with my LLM's but not being a python devs means the current tooling implementation is somewhat "out of reach".

I don't really want to have to deal with encapsulating a python framework and then building python functions in order to expose functionality to my LLM's so I set about figuring out how to prompt it into building out json blobs that represent HTTP requests.

I've had some success with this but I was saddened to see that I could not tell the LLM "this is the tool response" in any way over a http request to the host.

I was hoping to be able to just add a new message into the history then push that to the Ollama service like any other chat request then have it process that however it sees fit.

I see that other's have raised this sort of question before but I wondered if it was reasonable to suggest that passing a history item with a role of "tool" was a nice tidy fix for this situation.

this way in my agentic implementation wrapper around the LLM I don't have to concern myself with the inner workings of Ollama and instead can remain in the friendly http environment that's exposed.

Is this something that community would consider for accessibility for non python devs?
essentially I want to be able to explain to the model "here's an endpoint you can use to search for tools, and having founc the results you want, you can then call them based on their indivdual definition information".
Or perhaps stuff like "weather api exists here {some url} what's the weather in london today?" and assuming I meet some interface criteria for http endpoints "a standard for LLM tools for models" if you like, the back end can simply "wire this up" for me.

This would then allow me to focus on the tooling itself instead of the details of how to wire it up in a language i don't normally use.

Originally created by @TehWardy on GitHub (Aug 9, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/11830 I'm loving that Ollama spins me up a HTTP endpoint that I can use to interact with my LLM's but not being a python devs means the current tooling implementation is somewhat "out of reach". I don't really want to have to deal with encapsulating a python framework and then building python functions in order to expose functionality to my LLM's so I set about figuring out how to prompt it into building out json blobs that represent HTTP requests. I've had some success with this but I was saddened to see that I could not tell the LLM "this is the tool response" in any way over a http request to the host. I was hoping to be able to just add a new message into the history then push that to the Ollama service like any other chat request then have it process that however it sees fit. I see that other's have raised this sort of question before but I wondered if it was reasonable to suggest that passing a history item with a role of "tool" was a nice tidy fix for this situation. this way in my agentic implementation wrapper around the LLM I don't have to concern myself with the inner workings of Ollama and instead can remain in the friendly http environment that's exposed. Is this something that community would consider for accessibility for non python devs? essentially I want to be able to explain to the model "here's an endpoint you can use to search for tools, and having founc the results you want, you can then call them based on their indivdual definition information". Or perhaps stuff like "weather api exists here {some url} what's the weather in london today?" and assuming I meet some interface criteria for http endpoints "a standard for LLM tools for models" if you like, the back end can simply "wire this up" for me. This would then allow me to focus on the tooling itself instead of the details of how to wire it up in a language i don't normally use.
GiteaMirror added the feature request label 2026-04-12 20:01:00 -05:00
GiteaMirror commented 2026-04-12 20:01:01 -05:00
Author
Owner
Copy Link

@TehWardy commented on GitHub (Aug 9, 2025):

Ok this looks like a classic case of RTFM ... I'm gonna close this.

<!-- gh-comment-id:3170598438 --> @TehWardy commented on GitHub (Aug 9, 2025): Ok this looks like a classic case of RTFM ... I'm gonna close this.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#7851
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?