bb -m prompts --help
bb -m prompts
bb -m prompts /Users/slim/docker/labs-ai-tools-for-devs jimclark106 darwin prompts/docker
bb -m prompts /Users/slim/docker/labs-ai-tools-for-devs jimclark106 darwin prompts/docker --pretty-print-prompts
bb -m prompts --host-dir /Users/slim/docker/labs-ai-tools-for-devs \
--platform darwin \
--prompts-dir prompts/docker \
--pretty-print-prompts
bb -m prompts --host-dir /Users/slim/docker/labs-ai-tools-for-devs \
--platform darwin \
--prompts-dir prompts/project_type/ \
--pretty-print-prompts
Make sure the prompts/project_type prompts work on their own.
bb -m prompts run /Users/slim/docker/labs-make-runbook jimclark106 darwin prompts/project_type --debug
bb -m prompts run /Users/slim/docker/labs-make-runbook jimclark106 darwin prompts/project_type --nostream
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--platform darwin \
--prompts-dir prompts/project_type \
--nostream \
--model "llama3.1" \
--url http://localhost:11434/v1/chat/completions
TODO - this should fail better because the prompts-dir is not valid.
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--platform darwin \
--prompts-dir prompts \
--nostream \
--model "llama3.1" \
--url http://localhost:11434/v1/chat/completions
Now, verify that the prompts/dockerfiles prompts work with gpt-4
.
bb -m prompts run /Users/slim/docker/labs-make-runbook jimclark106 darwin prompts/dockerfiles
Now, let's do the same thing using gpt-4 but without streaming.
bb -m prompts run /Users/slim/docker/labs-make-runbook jimclark106 darwin prompts/dockerfiles --nostream
Now, let's try with llama3.1.
# docker:command=llama
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-dir prompts/dockerfiles_llama3.1 \
--url http://localhost:11434/v1/chat/completions \
--model "llama3.1" \
--nostream \
Now, let's try with mistral-nemo
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-dir prompts/dockerfiles_mistral-nemo \
--url http://localhost:11434/v1/chat/completions \
--model "mistral-nemo" \
--nostream \
Mistral is kind of doing function calls but not openai compatible ones. It's listing a set of functions to call and not getting the arguments correct.
bb -m prompts run /Users/slim/docker/labs-make-runbook jimclark106 darwin prompts/dockerfiles \
--url http://localhost:11434/v1/chat/completions \
--model "mistral:latest" \
--pretty-print-prompts
llama3-groq-tool-use:latest is writing functions but with a mix of xml and json markup. It's not compatible with openai currently. Also, the finish-reason is stop, instead of "tool_calls". So the conversation loop ends too.
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-dir prompts/dockerfiles \
--url http://localhost:11434/v1/chat/completions \
--model "llama3-groq-tool-use:latest"
rm ~/docker/labs-make-runbook/qrcode.png
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-file prompts/qrencode/README.md
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-file /Users/slim/docker/labs-ai-tools-for-devs/prompts/curl/README.md \
--debug
open ~/docker/labs-make-runbook/qrcode.png
bb -m prompts run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform darwin \
--prompts-file prompts/qrencode/README.md \
--url http://localhost:11434/v1/chat/completions \
--model "llama3.1" \
--nostream \
--debug
docker run
--rm \
--pull=always \
-it \
-v /var/run/docker.sock:/var/run/docker.sock \
--mount type=bind,source=$PWD,target=/app/local \
--workdir /app \
--mount type=volume,source=docker-prompts,target=/prompts \
--mount type=bind,source=$HOME/.openai-api-key,target=/root/.openai-api-key \
vonwig/prompts:local \
run \
--host-dir /Users/slim/docker/labs-make-runbook \
--user jimclark106 \
--platform "$(uname -o)" \
--prompts-dir local/prompts/dockerfiles
#docker:command=clean-local-images
bb -m clean-local-images
docker run --rm \
-it \
-v /var/run/docker.sock:/var/run/docker.sock \
--mount type=bind,source=$PWD,target=/app/local \
--workdir /app \
--mount type=volume,source=docker-prompts,target=/prompts \
--mount type=bind,source=$HOME/.openai-api-key,target=/root/.openai-api-key \
vonwig/prompts:local \
run \
/Users/slim/repo/labs-eslint-violations \
jimclark106 \
"$(uname -o)" \
local/prompts/eslint \
--pat "$(cat ~/.secrets/dockerhub-pat-ai-tools-for-devs.txt)" \
--thread-id "something"
- remove the openai key
- break the url with the --url flag
- choose a bad prompts dir
docker run --rm \
-it \
-v /var/run/docker.sock:/var/run/docker.sock \
--mount type=volume,source=docker-prompts,target=/prompts \
--mount type=bind,source=$PWD,target=/app/local \
--mount type=bind,source=$HOME/.openai-api-key,target=/root/.openai-api-key \
--workdir /app \
vonwig/prompts:local \
run \
--host-dir $PWD \
--user $USER \
--platform "$(uname -o)" \
--prompts-dir local/prompts/poem \