Salastil/odysseus
1
0
Fork 0
mirror of https://github.com/pewdiepie-archdaemon/odysseus.git synced 2026-06-28 23:52:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

feat(discovery): detect llama.cpp servers and label local providers (#4729)

Browse Source 
* feat(discovery): detect llama.cpp servers and label local providers

Scan port 8080 (llama-server) and 11435 (APFEL) during discovery, fingerprint
llama.cpp via its native /props endpoint, and label well-known local serving
ports (8080 llama.cpp, 8000 vLLM, 1234 LM Studio, 11434 Ollama) consistently
in both the Python provider helper and the JS endpoint UI. Adds a llama.cpp
hint to the /setup slash command.

* fix(discovery): don't infer the serving tool from the port alone

Per review: vLLM, SGLang, llama.cpp and plain OpenAI-compatible servers all
share 8000/8080, so labeling by port mislabels real setups (a vLLM box on 8080
shown as llama.cpp). Drop the port->tool assertions from _provider_label and
providerLabel; the authoritative signal is the /props fingerprint done during
discovery, which is unchanged. Loopback now reads a neutral 'local endpoint' /
'Local'. Tests updated to assert the neutral labels.
This commit is contained in:
Joel Alejandro Escareño Fernández
2026-06-23 23:39:56 +02:00
committed by GitHub
parent 72c0bde8a9
commit e0ccf250a4
9 changed files with 330 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
static/js/admin.js
+20 -3
View File
@@ -1603,8 +1603,8 @@ function initEndpointForm() {
wrap.style.cssText = 'display:flex;align-items:center;padding:8px 0;';
wrap.appendChild(wp.element);
const txt = document.createElement('span');
txt.textContent = 'Scanning ports 8000-8020 and 11434 for model servers...';
txt.style.cssText = 'opacity:0.7;';
txt.textContent = 'Scanning ports 8000-8020, 8080, 1234, 11434, and 11435 for model servers...';
txt.style.cssText = 'font-size:12px;opacity:0.7;';
wrap.appendChild(txt);
msg.appendChild(wrap);
discoverBtn._wp = wp;
@@ -1619,12 +1619,24 @@ function initEndpointForm() {
} else {
// Auto-add each discovered endpoint. Server dedupes on base_url
// and returns `existing: true` for already-registered ones.
// Map fingerprinted provider IDs to friendly display names.
const _PROVIDER_DISPLAY = {
llamacpp: 'llama.cpp', lmstudio: 'LM Studio', vllm: 'vLLM',
ollama: 'Ollama',
};
let added = 0;
let skipped = 0;
for (const item of items) {
const base = item.url.replace('/chat/completions', '').replace(/\/$/, '');
const providerDisplay = _PROVIDER_DISPLAY[item.provider] || null;
const fd = new FormData();
fd.append('base_url', base);
if (providerDisplay) {
// Use "Provider (host:port)" so the endpoint is immediately
// identifiable in the list, e.g. "llama.cpp (localhost:8080)".
const hostPart = base.replace(/^https?:\/\//, '').split('/')[0];
fd.append('name', `${providerDisplay} (${hostPart})`);
}
fd.append('endpoint_kind', 'local');
fd.append('model_refresh_mode', 'auto');
fd.append('skip_probe', 'false');
@@ -1638,7 +1650,12 @@ function initEndpointForm() {
}
}
const totalModels = items.reduce((n, i) => n + (i.models ? i.models.length : 0), 0);
const parts = [`Found ${items.length} server${items.length !== 1 ? 's' : ''} with ${totalModels} model${totalModels !== 1 ? 's' : ''}`];
const serverNames = items.map(i =>
(_PROVIDER_DISPLAY[i.provider] || i.url.replace(/^https?:\/\//, '').split('/')[0])
);
const parts = [
`Found ${items.length} server${items.length !== 1 ? 's' : ''} (${serverNames.join(', ')}) with ${totalModels} model${totalModels !== 1 ? 's' : ''}`,
];
if (added) parts.push(`added ${added} new`);
if (skipped) parts.push(`${skipped} already added`);
msg.innerHTML = parts.join(' — ');