Skip to main content

​
What Are Station Agents?

Station agents are AI-powered automation tools that:
  • Execute multi-step tasks autonomously
  • Access tools and APIs through MCP servers
  • Run on your infrastructure with your credentials
Unlike ChatGPT or Claude, Station agents have direct access to your tools (AWS, databases, filesystems) and can execute complex workflows without copy-pasting commands.

​
Dotprompt Format

Every agent is a .prompt file with YAML frontmatter and a template body: 
---
metadata:
  name: "Cost Analyzer"
  description: "Analyzes AWS costs and finds savings"
  tags: ["finops", "aws"]
model: gpt-4o-mini
max_steps: 8
tools:
  - "__get_cost_and_usage"
  - "__list_cost_allocation_tags"
---

{{role "system"}}
You are a FinOps analyst specializing in AWS cost analysis.

Your process:
1. Query AWS Cost Explorer for the past 30 days
2. Identify cost spikes exceeding 20% baseline
3. Break down costs by service
4. Provide actionable recommendations

{{role "user"}}
{{userInput}}

​
Frontmatter Reference

FieldTypeRequiredDescription
metadata.namestringYesAgent display name
metadata.descriptionstringYesWhat the agent does
metadata.tagsarrayNoCategories for search
modelstringYesAI model to use
max_stepsintegerYesMaximum tool calls (5-15 typical)
toolsarrayYesMCP tools the agent can use

​
Model Selection

ModelUse CaseSpeedCost
gpt-4o-miniSimple tasks, file opsFastLow
gpt-4oComplex reasoningMediumMedium
claude-3-5-sonnet-latestCode analysis, securityMediumMedium

​
Creating Agents

# Interactive creation
stn agent create

# With flags
stn agent create --name "my-agent" --model gpt-4o-mini

​
Running Agents

# Run an agent
stn agent run "Cost Analyzer" "Analyze last month's AWS spending"

# With real-time output
stn agent run "Cost Analyzer" "Find cost savings" --tail

​
Discovering Tools

List available MCP tools for your agents: 
# All tools in current environment
stn mcp tools

# Search for specific tools
stn mcp tools | grep aws

​
Coding Backends

Agents can be enhanced with coding capabilities that allow them to read, write, and modify files.

​
OpenCode Backend

Enable AI coding with filesystem access via OpenCode: 
---
metadata:
  name: "Code Assistant"
  description: "AI coding assistant"
model: openai/gpt-4o
coding:
  enabled: true
---

You are an expert software engineer. Use coding tools to implement features.
This gives the agent these tools:
ToolDescription
coding_openStart a coding session in a directory
codeExecute coding tasks (read/write/refactor)
coding_closeEnd the session
coding_commitCommit changes to git
coding_pushPush to remote repository
The transport mechanism (HTTP or NATS) is configured at the Station level, not per-agent. This makes agent definitions portable across different deployments. See OpenCode Backend for backend configuration.

​
Coding Frontmatter Options

FieldTypeRequiredDescription
coding.enabledbooleanYesEnable coding tools for this agent
coding.workspace_pathstringNoDefault workspace directory

OpenCode Backend

Full guide to OpenCode integration, HTTP vs NATS backends, and container mode

​
Sandbox Backend

For isolated code execution in containers (different from OpenCode), use the sandbox backend: 
---
model: openai/gpt-4o
sandbox:
  mode: code
  session: workflow
---

Sandbox Execution

Run code safely in isolated Docker containers

​
Notifications

Enable agents to send webhook notifications (ntfy, Slack, or any webhook endpoint): 
---
model: anthropic/claude-sonnet-4-20250514
notify: true
---
You are a monitoring agent. Send notifications when tasks complete or issues arise.
This gives the agent the notify tool with parameters:
ParameterDescription
messageNotification content (required)
titleSubject line (optional)
prioritymin, low, default, high, urgent
tagsEmoji tags like ["warning", "white_check_mark"]

Notifications

Full guide to webhook configuration and formats

​
Next Steps

Notifications

Send webhook notifications from agents

MCP Tools

Connect agents to external services

Multi-Agent Teams

Build hierarchical agent teams

Workflows

Orchestrate multi-step agent tasks

Evaluation

Test and score your agents