Frontmatter
| id | 9766 |
| title | Implement complete openAiCompatible provider architecture (Settings, Docs, Embeddings) |
| state | Closed |
| labels | enhancementaiagent-role:dev |
| assignees | tobiu |
| createdAt | Apr 7, 2026, 10:25 PM |
| updatedAt | Apr 7, 2026, 10:29 PM |
| githubUrl | https://github.com/neomjs/neo/issues/9766 |
| author | tobiu |
| commentsCount | 1 |
| parentIssue | null |
| subIssues | [] |
| subIssuesCompleted | 0 |
| subIssuesTotal | 0 |
| blockedBy | [] |
| blocking | [] |
| closedAt | Apr 7, 2026, 10:26 PM |
Implement complete openAiCompatible provider architecture (Settings, Docs, Embeddings)
Closedenhancementaiagent-role:dev
tobiu
Apr 7, 2026, 10:26 PM
Input from Antigravity (Gemini 3.1 Pro):
✦ Integration completed. Refactored config types, updated AI_QUICK_START syntax mappings, and structurally integrated MLX v1/embeddings mapping using standard ES6 mechanisms. Successfully pushed to local and remote origin.
tobiu closed this issue on Apr 7, 2026, 10:26 PM
tobiu assigned to @tobiu on Apr 7, 2026, 10:29 PM
The
openAiCompatibleengine must be formally documented and structurally supported across the entire stack, including semantic vector embeddings.Scope:
config.mjsandconfig.template.mjsto add clear definitions for supported provider strings without breaking JSDoc parsing (@type {String}with a manual enumeration comment).apiKeyto theopenAiCompatibleconfig map, gated firmly with a security warning to only use.envinjections..github/AI_QUICK_START.mdto advertise the MLX / MLX-Server mappings alongside Ollama.ai/mcp/server/memory-core/services/TextEmbeddingService.mjsto securely dispatch POST requests tov1/embeddingsusing optional chaining and ES6 syntax standards organically.