commands

#commands

docker compose up

to run the docker container

make build

to build the solution

make run MLX_BACKEND=1

to run the solution with MLX_BACKEND instead of

make empty

to run an empty container to install dependencies #package installs

docker exec backend /bin/bash

to enter the backend container

pip install PACKAGE

to install dependencies When installing packages,

conda env export >> linux_environment.yml

after installing to freeze installs. #frontend package installs

cd client & npm i PACKAGE

#running tests

cd /opt/geist && PYTHONPATH=/opt/geist pytest

in the backend container

#SQLAlchemy When adding classes to sql alchemy, take the following example EXAMPLE: import datetime from sqlalchemy import Column, Integer, String, ForeignKey, LargeBinary, DateTime, Boolean, ARRAY, DateTime from sqlalchemy.orm import relationship, Session from app.models.database.database import Base, Session from sqlalchemy.dialects.postgresql import insert

class AgentPreset(Base): """ Class used to represent a preset for an agent. """ tablename = 'agent_preset' agent_preset_id = Column(Integer, primary_key=True, autoincrement=True) #language model presets name = Column(String) version = Column(String) description = Column(String) max_tokens = Column(Integer) n = Column(Integer) temperature = Column(Integer) top_p = Column(Integer) frequency_penalty = Column(Integer) presence_penalty = Column(Integer) tags = Column(String) #memory presets working_context_length = Column(Integer) long_term_context_length = Column(Integer) agent_type = Column(String) #prompt presets prompt = Column(String) #interactive_only - is not an independent agent. interactive_only = Column(Boolean) # optional processing settings process_world = Column(Boolean) #restriction relationships restrictions = relationship("Restriction", back_populates="agent_preset") create_date = Column(DateTime) update_date = Column(DateTime)

classes are stored in app >> models >> database.

#Adding Models Models should be added to scripts/copy_weights.py as well as the associated agent implementation inheriting from agents/base_agent. (GPT4 Agent, llama_agent, etc.)

#preferences prefer minimal inline implementations over extra dependency imports. Core libraries are better than pypi packages.

#SDLC

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running

docker compose up -d

, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

#commands

docker compose up

to run the docker container

make build

to build the solution

make run MLX_BACKEND=1

to run the solution with MLX_BACKEND instead of

make empty

to run an empty container to install dependencies #package installs

docker exec backend /bin/bash

to enter the backend container

pip install PACKAGE

to install dependencies When installing packages,

conda env export >> linux_environment.yml

after installing to freeze installs. #frontend package installs

cd client & npm i PACKAGE

#running tests

cd /opt/geist && PYTHONPATH=/opt/geist pytest

in the backend container

classes are stored in app >> models >> database.

#Adding Models Models should be added to scripts/copy_weights.py as well as the associated agent implementation inheriting from agents/base_agent. (GPT4 Agent, llama_agent, etc.)

#preferences prefer minimal inline implementations over extra dependency imports. Core libraries are better than pypi packages.

#SDLC

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running

docker compose up -d

, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

commands

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running
`docker compose up -d`
, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

Related Skills

Nano Banana Pro

Markdown Converter

1password

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running
`docker compose up -d`
, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

commands

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running docker compose up -d, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

Related Skills

Nano Banana Pro

Markdown Converter

1password

first, create a plan for your feature in /plans

next, implement the plan, adding the backend data models, middle data models, service layer, routes, backend tests,

finally, test the solution by running docker compose up -d, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

finally, test the solution by running
`docker compose up -d`
, then verifying no error logs in the docker container, then doing a curl command to localhost:3000

finally, test the solution by running
`docker compose up -d`
, then verifying no error logs in the docker container, then doing a curl command to localhost:3000