Poolside

https://docs.poolside.ai
Admin
Poolside is a secure, enterprise-grade AI platform for software engineering that runs entirely...
Tokens:78,738
Snippets:1,187
Trust Score:9.8
Update:2 months ago
Show doc for...
Tokens:
Raw
### Install OpenRouter client library

Source: https://docs.poolside.ai/api/overview

Commands to install the OpenRouter SDK.

```bash
npm install @openrouter/sdk
```

```bash
pnpm add @openrouter/sdk
```

```bash
yarn add @openrouter/sdk
```

```bash
pip install openrouter
```

--------------------------------

### Install kubectl binary directly

Source: https://docs.poolside.ai/deployment/on-prem/install

Downloads the kubectl binary using wget and installs it to the system path.

```bash
sudo zypper install -y wget
wget https://dl.k8s.io/release/v1.35.3/bin/linux/<arch>/kubectl
chmod +x kubectl
sudo install -o root -g root -m 0755 kubectl /usr/local/bin/kubectl
```

--------------------------------

### Install Poolside CLI on Windows

Source: https://docs.poolside.ai/cli/install

Use PowerShell to download and execute the installation script on Windows systems.

```powershell
irm https://downloads.poolside.ai/pool/install.ps1 | iex
```

--------------------------------

### Install OpenAI client library

Source: https://docs.poolside.ai/api/overview

Commands to install the OpenAI client library via package managers.

```bash
pip install openai
```

```bash
npm install openai
```

--------------------------------

### Install kubectl

Source: https://docs.poolside.ai/deployment/on-prem/install

Installs the kubectl package using yum.

```bash
sudo yum install -y kubectl
```

--------------------------------

### Install NVIDIA driver

Source: https://docs.poolside.ai/deployment/on-prem/install

Installs the signed NVIDIA open driver package.

```bash
sudo zypper install nvidia-open-driver-G07-signed-cuda-kmp-default
```

--------------------------------

### Example .env File

Source: https://docs.poolside.ai/resources/use-cases/documenting-code

A sample environment configuration file for production settings.

```text
DB_HOST=mydb.example.com
DB_PORT=5432
DB_NAME=production_db
PORT=8080
JWT_SECRET=super-secret-jwt-key-change-in-production
SALT_ROUNDS=12
```

--------------------------------

### Install Ollama

Source: https://docs.poolside.ai/resources/run-model-locally

Executes the installation script for Ollama on supported systems.

```bash
curl -fsSL https://ollama.com/install.sh | sh
```

--------------------------------

### Start a Poolside Session

Source: https://docs.poolside.ai/get-started/quickstart

Navigate to your project directory and initialize the agent.

```bash
cd <project-path>
pool
```

--------------------------------

### Install Poolside CLI on macOS or Linux

Source: https://docs.poolside.ai/cli/install

Use curl or wget to download and execute the installation script on Unix-based systems.

```bash
curl -fsSL https://downloads.poolside.ai/pool/install.sh | sh
```

```bash
wget -qO- https://downloads.poolside.ai/pool/install.sh | sh
```

--------------------------------

### Start Poolside Workloads

Source: https://docs.poolside.ai/deployment/on-prem/server-maintenance

Starts the RKE2 server if inactive, followed by the Poolside workloads.

```bash
sudo systemctl start poolside-services
```

--------------------------------

### Configure Kubernetes repository and install kubectl

Source: https://docs.poolside.ai/deployment/on-prem/install

Adds the official Kubernetes repository to zypper and installs the kubectl binary.

```bash
cat <<EOF | sudo tee /etc/zypp/repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=https://pkgs.k8s.io/core:/stable:/v1.35/rpm/
enabled=1
gpgcheck=1
gpgkey=https://pkgs.k8s.io/core:/stable:/v1.35/rpm/repodata/repomd.xml.key
EOF

sudo zypper install -y kubectl
```

--------------------------------

### Start a new session

Source: https://docs.poolside.ai/cli/troubleshooting

Initializes a new interactive session to test connectivity.

```bash
pool
```

--------------------------------

### Install system dependencies via zypper

Source: https://docs.poolside.ai/deployment/on-prem/install

Installs essential packages including iptables, tar, unzip, skopeo, and jq.

```bash
sudo zypper install -y iptables tar unzip skopeo jq
```

--------------------------------

### Verify Pi Installation

Source: https://docs.poolside.ai/tools/pi

Check the installed version of the Pi CLI tool.

```bash
pi --version
```

--------------------------------

### Send a Prompt

Source: https://docs.poolside.ai/get-started/quickstart

Example prompt to test the agent's functionality.

```text
Summarize this project and suggest one small improvement.
```

--------------------------------

### Configure Netplan for Network Interface

Source: https://docs.poolside.ai/deployment/on-prem/relocation

Example Netplan configuration for a new network interface.

```yaml
network:
  version: 2
  ethernets:
    NM-<id>:
      renderer: NetworkManager
      match:
        name: "<new-interface>"
      addresses:
        - "<new-ip>/<prefix>"
      nameservers:
        addresses:
          - <new-dns-server-1>
          - <new-dns-server-2>
      dhcp6: true
      wakeonlan: true
      networkmanager:
        name: "Wired connection 4"
        passthrough:
          connection.autoconnect-priority: "-999"
          ethernet._: ""
          ipv4.address1: "<new-ip>/<prefix>,<new-gateway>"
          ipv4.method: "manual"
          ipv6.addr-gen-mode: "default"
          ipv6.ip6-privacy: "-1"
          proxy._: ""
```

--------------------------------

### Verify Ollama Installation

Source: https://docs.poolside.ai/resources/run-model-locally

Lists the currently available models to confirm that Ollama is correctly installed and running.

```bash
ollama list
```

--------------------------------

### Install the inference chart

Source: https://docs.poolside.ai/deployment/cloud/upstream-kubernetes/install

Deploys the inference chart to the poolside-models namespace using a values file.

```bash
helm install inference ./charts/inference \
  --namespace poolside-models \
  -f ./inference_values.yaml
```

```bash
helm install inference ./charts/inference \
  --namespace poolside-models \
  -f ./inference_values.yaml \
  --set-file s3.caBundle=<path-to-s3-ca.crt>
```

--------------------------------

### Install NVIDIA compute utilities

Source: https://docs.poolside.ai/deployment/on-prem/install

Installs matching compute utilities and persistence daemon based on the detected driver version.

```bash
sudo zypper install -y "nvidia-compute-utils-G07 = ${version}" "nvidia-persistenced = ${version}"
```

--------------------------------

### Example personal AGENTS.md

Source: https://docs.poolside.ai/agent-instructions

A template for defining response and coding preferences using Markdown.

```markdown
## Response preferences

- Put the code before the explanation.
- Keep explanations concise unless I ask for more detail.
- When you suggest shell commands, explain any destructive steps.

## Coding preferences

- Prefer immutable patterns over mutation where reasonable.
- Add docstrings to new public functions.
```

--------------------------------

### Launch Pi CLI

Source: https://docs.poolside.ai/tools/pi

Start the Pi application in your terminal.

```bash
pi
```

--------------------------------

### Start ACP server with environment variables

Source: https://docs.poolside.ai/cli/cli-reference

Starts the ACP server over standard input/output with specific environment variables.

```bash
KEY=VALUE pool acp
```

--------------------------------

### Preserve Installation State

Source: https://docs.poolside.ai/deployment/on-prem/upgrade

Copy the existing installation state directory to the new bundle to maintain configuration and certificates.

```bash
cp -aT "$OLD_BUNDLE/poolside-install" "$NEW_BUNDLE/poolside-install"
```

--------------------------------

### Configure Goose CLI

Source: https://docs.poolside.ai/tools/goose

Initial command to start the configuration wizard for provider settings.

```bash
goose configure
```

--------------------------------

### Add Kubernetes repository

Source: https://docs.poolside.ai/deployment/on-prem/install

Configures the yum repository for installing kubectl.

```bash
cat <<EOF | sudo tee /etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=https://pkgs.k8s.io/core:/stable:/v1.33/rpm/
enabled=1
gpgcheck=1
gpgkey=https://pkgs.k8s.io/core:/stable:/v1.33/rpm/repodata/repomd.xml.key
EOF
```

--------------------------------

### Install RKE2 via Terraform

Source: https://docs.poolside.ai/deployment/on-prem/install

Run these commands from the 01-infra-rke2 directory to initialize and apply the RKE2 infrastructure module.

```bash
sudo TF_CLI_CONFIG_FILE=<bundle-path>/poolside-terraform.tfrc /usr/local/bin/terraform init
sudo TF_CLI_CONFIG_FILE=<bundle-path>/poolside-terraform.tfrc /usr/local/bin/terraform apply
```

```bash
sudo /usr/local/bin/terraform init
sudo /usr/local/bin/terraform apply
```

--------------------------------

### Example Project AGENTS.md

Source: https://docs.poolside.ai/agent-instructions

Defines global project structure, development workflows, and coding conventions at the repository root.

```markdown
## Project overview

This repository contains the main application code. The `src/` directory contains product code, `tests/` contains automated tests, and `docs/` contains user-facing documentation.

## Development

- Check `README.md` for setup steps before installing dependencies.
- Run the relevant test command before finishing a code change.
- Use the existing formatter and linter configuration.

## Conventions

- Keep changes focused on the requested task.
- Add or update tests when behavior changes.
- Follow existing naming and file organization patterns.
```

--------------------------------

### Configure sandbox settings in YAML

Source: https://docs.poolside.ai/settings-file-reference

Example configuration for the sandbox block within the poolside settings file.

```yaml
sandbox:
  image: poolsideengineering/ubuntu-with-tools
  env_vars:
    NODE_ENV: development
  secrets:
    - GITHUB_TOKEN
  filesystem:
    workspaces:
      access: read-only
    mounts:
      - host: <absolute-host-path>
        sandbox: <absolute-sandbox-path>
        access: read-only
  network:
    policy: allow-list
    egress:
      allowed_domains:
        - poolside.ai
      allowed_cidrs:
        - 10.0.0.0/8
```

--------------------------------

### Start session with specific model

Source: https://docs.poolside.ai/cli/interactive-mode

Initiates a new CLI session using a designated model identifier.

```bash
pool --model poolside/laguna-m.1
```

--------------------------------

### Start Goose Session

Source: https://docs.poolside.ai/tools/goose

Command to launch an interactive session within a project directory.

```bash
goose session
```

--------------------------------

### Manage MCP servers and secrets via CLI

Source: https://docs.poolside.ai/cli/cli-reference

Examples for adding various types of MCP servers, passing environment variables or headers, and performing basic inspection and removal operations.

```bash
# Command-based server over stdio
pool mcp add filesystem -- node filesystem-server.js

# Remote HTTP server
pool mcp add --transport http notion https://mcp.notion.com/mcp

# Remote SSE server
pool mcp add --transport sse linear https://mcp.linear.app/sse

# Pass environment variables or HTTP headers
pool mcp add --env API_KEY=<api-key> myserver -- npx -y myserver-mcp
pool mcp add --transport http --header "Authorization: Bearer $TOKEN" svc https://example.com/mcp

# Inspect and remove
pool mcp list
pool mcp get <name>
pool mcp remove <name>
```

--------------------------------

### Execute Pool History Commands

Source: https://docs.poolside.ai/cli/cli-reference

Examples of common history commands for sessions, logs, and trajectories.

```bash
pool history sessions
pool history logs --latest --pretty
pool history trajectories --latest
```

--------------------------------

### Visualize AGENTS.md file hierarchy

Source: https://docs.poolside.ai/agent-instructions

A directory structure example showing how nested AGENTS.md files are organized within a repository.

```text
repo/
├── AGENTS.md              # instructions for the whole repository
├── app/
│   ├── AGENTS.md          # instructions for all app code
│   └── ui/
│       └── AGENTS.md      # instructions for UI code
└── api/
    └── AGENTS.md          # instructions for API code
```

--------------------------------

### Refactor authentication module

Source: https://docs.poolside.ai/resources/example-workflows

Use this prompt to guide the refactoring of an authentication module to a new library.

```text
@src/auth/ Refactor the authentication module to use the new OAuth2 library.
Update all imports and ensure tests still pass.
```

--------------------------------

### Connect to agent servers via CLI

Source: https://docs.poolside.ai/cli/other-agent-servers

Start the pool CLI with a specific agent server using the --agent-server flag.

```bash
# Claude Agent
pool --agent-server claude-agent-acp

# Codex
pool --agent-server codex-acp

# Gemini
pool --agent-server "gemini --acp"

# Remote ACP server
pool --agent-server http://localhost:3284/acp
```

--------------------------------

### Initialize settings file

Source: https://docs.poolside.ai/mcp-servers

Create the configuration directory and settings file if they do not exist.

```bash
mkdir -p ~/.config/poolside
touch ~/.config/poolside/settings.yaml
```

--------------------------------

### Retrieve installed NVIDIA driver version

Source: https://docs.poolside.ai/deployment/on-prem/install

Extracts the version string of the installed NVIDIA driver for use in subsequent package installations.

```bash
version=$(rpm -qa --queryformat '%{VERSION}\n' nvidia-open-driver-G07-signed-cuda-kmp-default | cut -d "_" -f1 | sort -u | tail -n 1)
```

--------------------------------

### Initialize a SKILL.md file

Source: https://docs.poolside.ai/skills

Create the required metadata file for a new skill.

```bash
touch .poolside/skills/code-review/SKILL.md
```

--------------------------------

### Example Directory AGENTS.md

Source: https://docs.poolside.ai/agent-instructions

Defines specific rules, constraints, and development workflows for a single directory within the repository.

```markdown
## Directory overview

This directory contains frontend code for the application UI.

## Development

- Run frontend tests before finishing changes in this directory.
- Use the existing component patterns.
- Do not edit generated files manually.

## Conventions

- Keep components small and focused.
- Follow existing naming and file organization patterns.
```

--------------------------------

### Directory structure for custom TLS certificates

Source: https://docs.poolside.ai/deployment/on-prem/install

Recommended file structure for BYO certificate files within the installation bundle.

```text
<bundle-path>/poolside-install/byo-certs/
├── ca.crt       # CA certificate (root, or root and intermediate chain)
├── server.crt   # Server certificate signed by the CA
└── server.key   # Server private key
```

--------------------------------

### Install Poolside Agent CLI

Source: https://docs.poolside.ai/get-started/quickstart

Commands to install the CLI on various operating systems.

```bash
curl -fsSL https://downloads.poolside.ai/pool/install.sh | sh
```

```bash
wget -qO- https://downloads.poolside.ai/pool/install.sh | sh
```

```powershell
irm https://downloads.poolside.ai/pool/install.ps1 | iex
```

--------------------------------

### Initialize ECR Repositories

Source: https://docs.poolside.ai/deployment/cloud/aws-eks/install

Creates necessary ECR repositories for the container images found in the local bundle.

```bash
for image_name in $(find ./containers -name "*.tar" -type f -exec basename {} .tar \; | sed 's/__.*//' | sort -u); do
  aws ecr describe-repositories --repository-names "$image_name" --region <aws-region> >/dev/null 2>&1 \
    || aws ecr create-repository --repository-name "$image_name" --region <aws-region>
done
```

--------------------------------

### Create a local skill directory

Source: https://docs.poolside.ai/skills

Initialize a new skill directory within the project's .poolside/skills path.

```bash
mkdir -p .poolside/skills/code-review
```

--------------------------------

### Start Poolside workloads

Source: https://docs.poolside.ai/deployment/on-prem/relocation

Start the Poolside services unit or re-run the inference phase Terraform scripts.

```bash
sudo systemctl start poolside-services
```

```bash
cd <bundle-path>/04-poolside-inference
terraform init
terraform apply
```

```bash
cd <bundle-path>/04-poolside-inference
TF_CLI_CONFIG_FILE=<bundle-path>/poolside-terraform.tfrc terraform init
TF_CLI_CONFIG_FILE=<bundle-path>/poolside-terraform.tfrc terraform apply
```

--------------------------------

### Initialize Inference Configuration

Source: https://docs.poolside.ai/deployment/cloud/upstream-kubernetes/install

Copies the default values file to the bundle root for environment-specific customization.

```bash
cp ./charts/inference/values.yaml ./inference_values.yaml
```

--------------------------------

### Install and Verify Poolside Agent CLI

Source: https://docs.poolside.ai/resources/run-model-locally

Installation script for the Poolside Agent CLI and command to verify the current version.

```bash
curl -fsSL https://downloads.poolside.ai/pool/install.sh | sh
```

```bash
pool --version
```

--------------------------------

### List zypper repositories

Source: https://docs.poolside.ai/deployment/on-prem/install

Displays all configured zypper repositories to verify the NVIDIA-GPU-Compute-Toolkit-CUDA source.

```bash
sudo zypper lr
```

--------------------------------

### Apply Netplan Configuration

Source: https://docs.poolside.ai/deployment/on-prem/relocation

Commands to test and apply new network configurations.

```bash
sudo netplan try
sudo netplan apply
```

--------------------------------

### Define skill metadata and instructions

Source: https://docs.poolside.ai/skills

Configure the skill's frontmatter and documentation within SKILL.md.

```markdown
---
name: code-review
description: Review code changes against team review guidelines. Use when reviewing pull requests, diffs, or local code changes.
---

# Code review

Review code changes against the team's review checklist.
```

--------------------------------

### Navigate to Project Directory

Source: https://docs.poolside.ai/tools/ollama

Change the current working directory to the target project path.

```bash
cd <project-path>
```

--------------------------------

### GET /v1/models

Source: https://docs.poolside.ai/api/openai-api-examples

Retrieves a list of all available models and their corresponding IDs.

```APIDOC
## GET /v1/models

### Description
Retrieves a list of all available models and their IDs from the Poolside inference service.

### Method
GET

### Endpoint
https://inference.poolside.ai/v1/models

### Parameters
#### Headers
- **Authorization** (string) - Required - Bearer <api-key>
- **Accept** (string) - Required - application/json, application/problem+json

### Response
#### Success Response (200)
- **data** (array) - List of model objects
- **object** (string) - The type of object returned

#### Response Example
{
  "data": [
    {
      "id": "poolside/laguna-m.1",
      "created": 1751637312,
      "owned_by": "system",
      "object": "model"
    }
  ],
  "object": "list"
}
```

--------------------------------

### Check Model Initialization Logs

Source: https://docs.poolside.ai/deployment/on-prem/upgrade

Examine logs from the model-downloader container to troubleshoot initialization or loading issues.

```bash
kubectl logs <pod-name> -c model-downloader -n poolside-models
```

--------------------------------

### Set Provider ID

Source: https://docs.poolside.ai/tools/opencode

Specify the provider identifier during the connection setup.

```text
poolside
```

--------------------------------

### Configure JetBrains via CLI

Source: https://docs.poolside.ai/tools/jetbrains

Run this command to automatically generate the necessary configuration file at ~/.jetbrains/acp.json.

```bash
pool acp setup --editor jetbrains
```

--------------------------------

### pool acp setup

Source: https://docs.poolside.ai/cli/cli-reference

Configures supported editors to use Poolside.

```APIDOC
## pool acp setup

### Description
Configures Zed or JetBrains to use Poolside.

### Parameters
#### Flags
- **--editor** (string) - Required - Editor to configure. Supported values: `zed`, `jetbrains`.
```

--------------------------------

### Verify Poolside CLI Version

Source: https://docs.poolside.ai/cli/install

Check the installed version of the Poolside CLI.

```bash
pool --version
```

--------------------------------

### GET /v1/models

Source: https://docs.poolside.ai/deployment/cloud/openshift/install

Retrieves a list of available models served by the inference API.

```APIDOC
## GET /v1/models

### Description
Retrieves a list of available models served by the inference API.

### Method
GET

### Endpoint
https://<route-host>/v1/models
```

--------------------------------

### Troubleshoot GPU and model loading issues

Source: https://docs.poolside.ai/deployment/on-prem/install

Commands to diagnose GPU availability and inspect model pod logs.

```bash
lspci | grep -i nvidia

```

```bash
kubectl get nodes -o jsonpath='{range .items[*]}{.metadata.name}{" "}{.status.allocatable.nvidia\.com/gpu}{"\n"}{end}'

```

```bash
kubectl get pods -n poolside-models

```

```bash
kubectl logs <pod-name> -c model-downloader -n poolside-models

```

--------------------------------

### Call the inference API

Source: https://docs.poolside.ai/deployment/cloud/upstream-kubernetes/install

Examples of sending chat completion requests to the model endpoint.

```bash
curl http://<model-hostname>/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<served-model-name>",
    "messages": [{"role": "user", "content": "Write a function that reverses a string."}]
  }'
```

```bash
curl http://laguna-m.poolside.local/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Laguna",
    "messages": [{"role": "user", "content": "Write a function that reverses a string."}]
  }'
```

```bash
curl http://<model-hostname>/v1/chat/completions \
  -H "Authorization: Bearer <vllm-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<served-model-name>",
    "messages": [{"role": "user", "content": "Write a function that reverses a string."}]
  }'
```

--------------------------------

### Interactive agent server selection

Source: https://docs.poolside.ai/cli/other-agent-servers

Launch the interactive menu to select from configured agent servers.

```bash
pool -s
```

--------------------------------

### Configure inference_values.yaml

Source: https://docs.poolside.ai/deployment/cloud/openshift/manage-models

Example configuration for updating the model path in the Helm values file.

```yaml
models:
  laguna:
    model: s3://<bucket-name>/checkpoints/laguna-<version>
    modelName: Laguna
    modelType: agent
    gpus: 4
    routeHost: ""
```

--------------------------------

### Configure OpenAI SDK for Poolside

Source: https://docs.poolside.ai/api/overview

Initialize the OpenAI client with the Poolside base URL and API key.

```python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["POOLSIDE_API_KEY"],
    base_url="https://inference.poolside.ai/v1",
)

completion = client.chat.completions.create(
    model="poolside/laguna-m.1",
    messages=[{"role": "user", "content": "Hello Laguna"}],
)

print(completion.choices[0].message.content)
```

```typescript
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.POOLSIDE_API_KEY,
  baseURL: "https://inference.poolside.ai/v1",
});

const completion = await client.chat.completions.create({
  model: "poolside/laguna-m.1",
  messages: [{ role: "user", content: "Hello Laguna" }],
});

console.log(completion.choices[0].message.content);
```

--------------------------------

### Use OpenRouter SDK

Source: https://docs.poolside.ai/api/overview

Initialize the OpenRouter client and send a chat request.

```python
import os
from openrouter import OpenRouter

with OpenRouter(api_key=os.environ["OPENROUTER_API_KEY"]) as client:
    response = client.chat.send(
        model="poolside/laguna-m.1",
        messages=[{"role": "user", "content": "Hello Laguna"}],
    )

    print(response.choices[0].message.content)
```

```typescript
import { OpenRouter } from "@openrouter/sdk";

const client = new OpenRouter({
  apiKey: process.env.OPENROUTER_API_KEY,
});

const completion = await client.chat.send({
  chatRequest: {
    model: "poolside/laguna-m.1",
    messages: [{ role: "user", content: "Hello Laguna" }],
  },
});

console.log(completion.choices[0].message.content);
```

--------------------------------

### Create a basic Mermaid flowchart

Source: https://docs.poolside.ai/resources/use-cases/visualizing-data-flows

A simple example of a left-to-right flowchart using nodes and arrows.

```mermaid
graph LR
    A[Start] --> B[Process] --> C[End]
```

--------------------------------

### Check host architecture

Source: https://docs.poolside.ai/deployment/on-prem/install

Verify the host architecture to determine whether to download amd64 or arm64 binaries.

```bash
uname -m
```

--------------------------------

### Configure Model Inference Variables

Source: https://docs.poolside.ai/deployment/on-prem/install

Example configuration for defining model deployment details in terraform.tfvars.

```hcl
deployment_name = "poolside-server"

models = {
  agent = {
    s3_uri = "s3://poolside-models/<model-checkpoint-name>"
    ingress_host_name = "poolside-models-agent.poolside.local"
    gpus = 1
    replicas = 1
    model_type = "agent"
  }
}
```

--------------------------------

### Verify model deployment

Source: https://docs.poolside.ai/deployment/cloud/openshift/manage-models

Check the status of the pods and routes to ensure the new model is running correctly.

```bash
oc get pods -n poolside-models
oc get route inference-<model-key> -n poolside-models
```

--------------------------------

### Configure inference_values.yaml

Source: https://docs.poolside.ai/deployment/cloud/openshift/install

Example configuration for deploying models with specific GPU requirements and OpenShift routes.

```yaml
image:
  # -- Registry you uploaded the atlas image to (required)
  registry: "<registry-host>"
  # -- Image name and tag come pre-set in the bundle to match the shipped image
  name: "atlas"
  tag: "<atlas-tag>"
# -- Name of the image pull secret for private registries (omit if your registry is public)
imagePullSecret: "poolside-registry-secret"
podSecurityContext:
  # -- Require non-root user. Do not set runAsUser on OpenShift; the SCC injects a UID from the namespace range.
  runAsNonRoot: true
  seccompProfile:
    # -- Seccomp profile type
    type: RuntimeDefault
s3:
  # -- Name of secret containing AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
  secretName: "aws-credentials"
  # -- Custom CA certificate bundle for S3 (required for NooBaa with the OpenShift service CA)
  caBundle: ""
authentication:
  # -- Name of secret containing VLLM_API_KEY for vLLM server authentication (set to "vllm-auth" if you created the optional secret in Step 4; leave empty to disable)
  secretName: ""
route:
  # -- Create a Route for every model
  enabled: true
  tls:
    # -- Terminate TLS at the OpenShift router
    enabled: true
    termination: edge
    insecureEdgeTerminationPolicy: Redirect
models:
  laguna-m:
    model: s3://<bucket-name>/checkpoints/laguna-m
    modelName: Lagunam
    modelType: agent
    gpus: 4
    # -- Route host for this model (leave empty for a router-generated hostname)
    routeHost: ""
  laguna-xs:
    model: s3://<bucket-name>/checkpoints/laguna-xs
    modelName: Lagunaxs
    modelType: agent
    gpus: 1
    # -- Route host for this model (leave empty for a router-generated hostname)
    routeHost: ""
```

--------------------------------

### Streaming response format

Source: https://docs.poolside.ai/api/openai-api-examples

Example of the server-sent events (SSE) format returned when streaming is enabled.

```text
data: {"model":"poolside/laguna-m.1","created":1754035552,"object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":""}]}

data: {"model":"poolside/laguna-m.1","created":1754035552,"object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"cURL","role":"assistant"},"finish_reason":""}]}

data: {"model":"poolside/laguna-m.1","created":1754035552,"object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":" ","role":"assistant"},"finish_reason":""}]}

data: {"model":"poolside/laguna-m.1","created":1754035552,"object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"is a","role":"assistant"},"finish_reason":""}]}

...
```

--------------------------------

### Configure reasoning effort

Source: https://docs.poolside.ai/api/openai-api-examples

Example request body including the reasoning object for OpenRouter-compatible models.

```json
{
  "model": "poolside/laguna-m.1",
  "messages": [
    {
      "content": "Explain cURL",
      "role": "user"
    }
  ],
  "reasoning": {
    "effort": "high"
  }
}
```

--------------------------------

### List Available Models

Source: https://docs.poolside.ai/tools/opencode

Refresh and display the available models after configuration.

```text
/models
```

--------------------------------

### Chat completion response structure

Source: https://docs.poolside.ai/api/openai-api-examples

Example of the JSON response returned by the chat completions API.

```json
{
  "model": "poolside/laguna-m.1",
  "created": 1751993576,
  "object": "chat.completion",
  "choices": [
    {
      "index": 0,
      "message": {
        "content": "cURL is a powerful command-line tool used for transferring data to or from a server, supporting various protocols such as HTTP, HTTPS, FTP, and more. It's widely used for testing APIs, downloading files, and automating HTTP requests. cURL allows you to specify headers, methods (GET, POST, PUT, DELETE, etc.), and data payloads, making it versatile for a range of web-related tasks.\n",
        "role": "assistant"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "completion_tokens": 88,
    "prompt_tokens": 447,
    "total_tokens": 535
  }
}
```

--------------------------------

### Set Bundle Path Variables

Source: https://docs.poolside.ai/deployment/on-prem/upgrade

Define shell variables for the current and new installation bundle directories.

```bash
export OLD_BUNDLE="<current-bundle-path>"
export NEW_BUNDLE="<new-bundle-path>"
```

--------------------------------

### pool acp

Source: https://docs.poolside.ai/cli/cli-reference

Starts the Poolside Agent Client Protocol (ACP) server over standard input and output.

```APIDOC
## pool acp

### Description
Starts the Poolside Agent Client Protocol (ACP) server over standard input and standard output.

### Parameters
#### Flags
- **--sandbox** (string) - Optional - Override sandbox usage. Supported values: `required`, `disabled`.
- **--settings** (string) - Optional - Apply extra settings via path to a YAML file or inline YAML content.
- **--version, -v** (flag) - Optional - Show the current version and exit.
```

--------------------------------

### Inspect model downloader logs

Source: https://docs.poolside.ai/deployment/cloud/aws-eks/install

Check init container logs to debug pods stuck in Init or restart loops.

```bash
kubectl logs -n poolside-models <pod-name> -c model-downloader
```
Tokens:
Raw