drwily/gogobee
1
0
Fork 0
mirror of https://github.com/prosolis/gogobee.git synced 2026-09-14 10:51:09 +00:00
Code Issues Packages Projects Releases Wiki Activity

llm: route every LLM caller through a shared backend client

Browse Source 
Replaces the per-plugin Ollama HTTP calls with internal/llm, which picks a
backend from the environment (vLLM or Ollama) behind one Chat interface, plus
internal/plugin/llm_client.go as the plugin-facing wrapper.

Startup now logs llm_backend/llm_endpoint/llm_model instead of the two
OLLAMA_* vars, which no longer describe where inference actually goes.

These files were already running in prod from the vLLM migration but had never
been committed; this is that live state, byte-for-byte.
This commit is contained in:
prosolis
2026-07-26 10:16:33 -07:00
parent 3f9c338e67
commit 583616f9d0
17 changed files with 603 additions and 434 deletions
Download Patch File Download Diff File Expand all files Collapse all files
internal/plugin/vibe.go
+4 -8
View File
@@ -120,9 +120,7 @@ func (p *VibePlugin) resetCooldown(roomID id.RoomID) {
}
func (p *VibePlugin) handleVibe(ctx MessageContext) error {
ollamaHost := os.Getenv("OLLAMA_HOST")
ollamaModel := os.Getenv("OLLAMA_MODEL")
if ollamaHost == "" || ollamaModel == "" {
if !llmConfigured() {
return p.SendReply(ctx.RoomID, ctx.EventID, "LLM is not configured.")
}
@@ -154,7 +152,7 @@ Describe the room's current vibe:`, botName, transcript)
slog.Error("vibe: send thinking", "err", err)
}
response, err := callOllama(ollamaHost, ollamaModel, prompt)
response, err := callLLM(prompt)
if err != nil {
slog.Error("vibe: ollama call", "err", err)
p.resetCooldown(ctx.RoomID) // Don't consume cooldown on failure
@@ -165,9 +163,7 @@ Describe the room's current vibe:`, botName, transcript)
}
func (p *VibePlugin) handleTLDR(ctx MessageContext) error {
ollamaHost := os.Getenv("OLLAMA_HOST")
ollamaModel := os.Getenv("OLLAMA_MODEL")
if ollamaHost == "" || ollamaModel == "" {
if !llmConfigured() {
return p.SendReply(ctx.RoomID, ctx.EventID, "LLM is not configured.")
}
@@ -199,7 +195,7 @@ Summary:`, tldrBotName, transcript)
slog.Error("vibe: send thinking", "err", err)
}
response, err := callOllama(ollamaHost, ollamaModel, prompt)
response, err := callLLM(prompt)
if err != nil {
slog.Error("vibe: ollama call", "err", err)
p.resetCooldown(ctx.RoomID) // Don't consume cooldown on failure