You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/agents/index.md
+43-21Lines changed: 43 additions & 21 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -18,6 +18,41 @@ VibePod manages each agent as a Docker container. Credentials and config are per
18
18
19
19
Start any agent for the first time with `vp run <agent>`. The container will prompt you to authenticate (browser OAuth, API key entry, or device flow depending on the provider). Once authenticated, credentials are written to the persisted config directory and reused on subsequent runs.
20
20
21
+
## Auto-pulling the latest image
22
+
23
+
VibePod automatically pulls the latest image for an agent before every run. This ensures you always start with the most up-to-date container without manual intervention.
24
+
25
+
To disable auto-pull globally:
26
+
27
+
```yaml
28
+
auto_pull: false
29
+
```
30
+
31
+
Or for a specific agent only:
32
+
33
+
```yaml
34
+
agents:
35
+
devstral:
36
+
auto_pull: false
37
+
```
38
+
39
+
Per-agent `auto_pull` takes precedence over the global setting. For example, you can disable it globally but keep it on for a specific agent:
40
+
41
+
```yaml
42
+
auto_pull: false # skip pull by default
43
+
agents:
44
+
claude:
45
+
auto_pull: true # except claude — always pull
46
+
```
47
+
48
+
You can also force a one-off pull via the CLI flag regardless of config:
49
+
50
+
```bash
51
+
vp run claude --pull
52
+
```
53
+
54
+
The resolution order is: `--pull`flag > per-agent `auto_pull` > global `auto_pull`.
55
+
21
56
## Overriding the image
22
57
23
58
You can point VibePod at a custom image via an environment variable:
@@ -132,7 +167,7 @@ The `init` commands run on every `vp run` for that agent and must be idempotent.
132
167
133
168
## Detached mode
134
169
135
-
Use `-d` / `--detach` to start an agent container in the background without attaching your terminal. This is useful when you want to customise the container environment before launching the agent interactively.
170
+
Use `-d` / `--detach` to start an agent container in the background without attaching your terminal. The agent process starts immediately inside the container — `-d` only controls whether VibePod attaches your terminal to it.
136
171
137
172
### Basic usage
138
173
@@ -147,31 +182,18 @@ The command prints the container name and returns immediately. You can also find
147
182
vp list --running
148
183
```
149
184
150
-
### Customizing the container before starting the agent
185
+
### Interacting with a detached container
151
186
152
-
A common workflow is to start detached, exec into the container to make adjustments, and then start the agent manually:
187
+
The agent is already running inside the container. You can exec into it to inspect state, install extra tools, or interact with the agent alongside its running process:
153
188
154
-
1. **Start the container in detached mode.**
155
-
156
-
```bash
157
-
vp run claude -d
158
-
# ✓ Started vibepod-claude-a1b2c3d4
159
-
```
160
-
161
-
2. **Exec into the running container.**
162
-
163
-
Use the container name printed above (or grab it from `vp list`):
164
-
165
-
```bash
166
-
docker exec -it vibepod-claude-a1b2c3d4 bash
167
-
```
168
-
169
-
3. **Apply your customizations** — install packages, edit config files, set environment variables, etc.
189
+
```bash
190
+
docker exec -it vibepod-claude-a1b2c3d4 bash
191
+
```
170
192
171
-
4. **Start the agent process** from inside the container when you are ready.
193
+
Use the container name printed by `vp run -d` or shown in `vp list`.
172
194
173
195
!!! tip
174
-
If you find yourself running the same setup steps every time, consider using [`agents.<agent>.init`](#init-scripts-before-startup) commands or [extending the base image](#image-customization-workflows) instead.
196
+
If you need to run setup commands **before** the agent launches, use [`agents.<agent>.init`](#init-scripts-before-startup) or [extend the base image](#image-customization-workflows) instead — these run inside the container before the agent process starts.
0 commit comments