You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Purpose and scope: Represent all the AI tooling on chain
Primary use cases: When something is not just "connected with AI", but is really focused on AI
Example entities/providers to include: Mosaia, Laika-AI
Columns (list):
slug, provider, actionButtons, toolType, planType, planName, price, trial, starred - like everywhere else
toolType enum is all the other categories that are there plus theoretically additional ones. The reasoning is that AI exists to support with specific areas, or do something instead of human, so I think it should follow the logic of category naming
deliveryType (same as in [DBIP] Extract few categories from DevTools - Testing and debugging #441 - not sure really about proper naming) - enum of CLI | Library | Desktop app | SaaS | Plugin
tag, description - same as in devTool.csv/services.csv
executionEnvironment - optional, may be used to specify if it may run only in specific envs. Doesn't matter for many AIs, so default may be "Any"
onChainFeatures (again, naming is rather draft) - JSON array field that answers if model may run anything on-chain (like, "send transactions", "create smart contracts", etc)
modelSourcing - hosted/self_hosted/both - whether tool provides model for you, or you should bring your own
Proposal type
Add category (table)
Affected scope (files/folders/chains)
ai.csv
Motivation / problem statement
AI is a most modern and hype narrative, a lot of devs use them. I think we need to extract AI from DevTools and make them a separate category.
Detailed proposal
Resolves #434 RFDBIP
If proposing a new/changed category (table):
slug, provider, actionButtons, toolType, planType, planName, price, trial, starred - like everywhere else
toolType enum is all the other categories that are there plus theoretically additional ones. The reasoning is that AI exists to support with specific areas, or do something instead of human, so I think it should follow the logic of category naming
deliveryType (same as in [DBIP] Extract few categories from DevTools - Testing and debugging #441 - not sure really about proper naming) - enum of CLI | Library | Desktop app | SaaS | Plugin
tag, description - same as in devTool.csv/services.csv
executionEnvironment - optional, may be used to specify if it may run only in specific envs. Doesn't matter for many AIs, so default may be "Any"
onChainFeatures (again, naming is rather draft) - JSON array field that answers if model may run anything on-chain (like, "send transactions", "create smart contracts", etc)
modelSourcing - hosted/self_hosted/both - whether tool provides model for you, or you should bring your own
Contact (optional)
No response
Rewards address (optional)
No response