dfdbaf55e94d9173f6710b236f905f72bc862075

 1diff --git a/README.md b/README.md
 2index eeb2ee19747033d480c083708fcd4f4e454652f8..6ecef5788837fb44efffc41528335fbc4fb04ab4 100644
 3--- a/README.md
 4+++ b/README.md
 5@@ -8,7 +8,7 @@ registers them as the `llama-cpp` provider in pi.
 6 **From the shell:**
 7 
 8 ```bash
 9-pi install https://github.com/huggingface/pi-llama
10+pi install git:github.com/huggingface/pi-llama
11 ```
12 
13 This clones to `~/.pi/agent/packages/pi-llama/` and adds an entry to your pi
14@@ -17,7 +17,7 @@ settings. Every future `pi` invocation auto-loads it.
15 **From inside an interactive pi session:**
16 
17 ```
18-!pi install https://github.com/huggingface/pi-llama
19+!pi install git:github.com/huggingface/pi-llama
20 ```
21 
22 Then run `/reload` (or restart pi) to load the extension.
23@@ -39,11 +39,11 @@ developing.
24 curl -LsSf https://llama.app/install.sh | bash
25 
26 # 2. Start it
27-llama-server
28+llama serve
29 
30-# 3. Launch pi
31+# 3. Launch pi in another terminal
32 pi
33 
34 # 4. Inside pi
35-/model              # pick llama-cpp/<id>
36+/model              # search "llama-cpp" to browse your local models
37 ```