Skip to content

feat(ai): add AIGROW memory provider preset - #516

Open
guglielmo-io wants to merge 1 commit into
tickernelz:mainfrom
guglielmo-io:feat/aigrow-memory-preset
Open

guglielmo-io wants to merge 1 commit into
tickernelz:mainfrom
guglielmo-io:feat/aigrow-memory-preset

Conversation

@guglielmo-io

Copy link
Copy Markdown

Summary

Adds an optional AIGROW memory provider preset on the existing OpenAI Chat Completions transport, following the merged UberLLM preset in #505 (and Standard Compute in #502). AIGROW sells genuine Claude, GPT, Gemini, Grok, Qwen and Kimi models through one OpenAI-compatible API on prepaid credit, at a reduced price; model ids are bare (claude-haiku-5.5, gpt-6.2-sol, ...). Each model can be checked against the maker's own endpoint: https://aigrow.gg/ai-api/genuine-models

  • memoryProvider: "aigrow": endpoint default https://aigrow.gg/v1, model default claude-haiku-5.5, AIGROW_API_KEY fallback, explicit env:// key configuration, and a distinct aigrow session tag.
  • Registered in memory-provider-presets.ts, so key-only configs resolve model and endpoint and readiness works the same as the other presets. Factory, types, config template, example config and all README languages updated.
  • Tests mirror the UberLLM ones: authentication, request serialization (tool_choice: "required"), tool response extraction, overrides, environment-key resolution, and rejection of remote-provider changes from project configuration.
  • No defaults change for existing users. Users opt in with their own key (https://aigrow.gg/dashboard/api, prepaid credit; guide at https://aigrow.gg/ai-api/guide). The docs name the destination for conversation data (https://aigrow.gg). Per-model prices: GET https://aigrow.gg/v1/models.

Test plan

  • Exercised the change locally: ran AigrowProvider.executeToolCall against the live API with a config built by buildMemoryProviderConfig({ memoryProvider: "aigrow", memoryApiKey }) and a forced save_memories tool. The default model claude-haiku-5.5 returned success: true with the tool arguments parsed ({"memories":["This repo uses pnpm as the package manager; never use npm.","Tests in this repo run with vitest."]}); gpt-6.2-sol and gemini-3.8-flash as memoryModel did the same.
  • Forced tool calls with the default memoryTemperature: 0.3 work on the default model, so no forceToolChoice: false is needed for it.

Checklist

  • Branch is based on main
  • bun test passes: full suite after bun run build, 1259 pass, 4 skip, 0 fail (163 files)
  • bun run typecheck passes
  • bun run check passes: prettier --check clean, eslint --max-warnings=0 clean, tsc --noEmit clean
  • Docs / README updated when behavior or contribution process changes

I work on AIGROW. AIGROW is sold by PeakCore LLC; the model makers did not review or endorse it.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant