Traycer: Spec-Driven Development

https://docs.traycer.ai/
Admin
Traycer AI enables Spec-Driven Development by transforming intent into structured plans,...
Tokens:20,905
Snippets:378
Trust Score:8
Update:3 months ago
Show doc for...
Tokens:
Raw
### Install Traycer CLI

Source: https://docs.traycer.ai/install

Install the standalone CLI tool using npm or Homebrew.

```bash
npm install -g @traycerai/cli
```

```bash
brew install traycerai/traycer/traycer
```

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

### Launch Traycer from WSL

Source: https://docs.traycer.ai/install

Execute the binary directly after a .deb or .rpm installation.

```bash
/opt/Traycer/traycer-desktop
```

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

### Verify CLI Installation

Source: https://docs.traycer.ai/cli/commands

Check the installed version of the Traycer CLI.

```bash
traycer --version
```

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

### Install Traycer Desktop via Homebrew

Source: https://docs.traycer.ai/install

Use Homebrew to install the Traycer desktop application on macOS.

```bash
brew install --cask traycerai/traycer/traycer-desktop
```

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

### Define Workspace-Level Agent Selection Guide

Source: https://docs.traycer.ai/settings/agents

Path for creating a workspace-specific agent selection guide to override global settings.

```text
<workspace>/.traycer/agent-selection-guide.md
```

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

### Install Traycer on Debian or Ubuntu via WSL

Source: https://docs.traycer.ai/install

Use these commands to download and install the .deb package on Debian-based Linux distributions.

```bash
wget https://github.com/traycerai/traycer/releases/latest/download/traycer-desktop-linux-amd64.deb
sudo apt install ./traycer-desktop-linux-amd64.deb
```

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

### Install Traycer via AppImage

Source: https://docs.traycer.ai/install

Download the portable AppImage, make it executable, and run it directly.

```bash
wget https://github.com/traycerai/traycer/releases/latest/download/traycer-desktop-linux-x86_64.AppImage
chmod +x traycer-desktop-linux-x86_64.AppImage
./traycer-desktop-linux-x86_64.AppImage
```

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

### Install Traycer on Fedora or RHEL via WSL

Source: https://docs.traycer.ai/install

Use these commands to download and install the .rpm package on Fedora or RHEL-based Linux distributions.

```bash
wget https://github.com/traycerai/traycer/releases/latest/download/traycer-desktop-linux-x86_64.rpm
sudo dnf install ./traycer-desktop-linux-x86_64.rpm
```

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

### Configuring Cursor API Key

Source: https://docs.traycer.ai/reference/troubleshooting

Environment variable configuration for Cursor provider authentication.

```bash
export CURSOR_API_KEY
```

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

### Cursor CLI Basic Usage

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Pipe the prompt into the cursor-agent.

```bash
#!/bin/sh
echo $TRAYCER_PROMPT | cursor-agent
```

```bash
#!/bin/sh
echo $TRAYCER_PROMPT | cursor-agent
```

```powershell
echo "$env:TRAYCER_PROMPT" | cursor-agent
```

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

### Reference Arguments in Command File

Source: https://docs.traycer.ai/extension/tasks/workflows

Use positional placeholders like $1 and $2 within command instructions to dynamically inject user-provided context.

```markdown
<processing_user_request>
The user wants to implement: $1
Using technology: $2
</processing_user_request>
```

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

### Factory Droid CLI Basic Usage

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Run Factory Droid with a file-based prompt.

```bash
#!/bin/sh
droid --file "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
droid --file "$TRAYCER_PROMPT"
```

```powershell
droid --file "$env:TRAYCER_PROMPT"
```

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

### Cline CLI Basic Usage

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Run the basic Cline CLI command with the provided prompt.

```bash
#!/bin/sh
cline "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
cline "$TRAYCER_PROMPT"
```

```powershell
cline "$env:TRAYCER_PROMPT"
```

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

### Execute CLI Agent with Prompt

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Basic execution of an agent using the TRAYCER_PROMPT environment variable.

```bash
#!/bin/sh
aider --yes-always "$TRAYCER_PROMPT"
```

```powershell
aider --yes-always "$env:TRAYCER_PROMPT"
```

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

### Call Command with Arguments

Source: https://docs.traycer.ai/extension/tasks/workflows

Execute a command by appending the required argument values directly after the command name.

```text
/create-feature User authentication OAuth2 and JWT
```

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

### Define a Handlebars template for coding agents

Source: https://docs.traycer.ai/extension/integrations/templates

Use frontmatter to specify the template display name and applicable context, then use {{planMarkdown}} to inject generated content.

```handlebars
---
displayName: Auto-Test Plan Template
applicableFor: plan
---

Follow the below plan verbatim. Trust the files and references. Do not re-verify
what's written in the plan.

{{planMarkdown}}

## After Implementation - Always Run Tests

After implementing all the changes above:

1. Run `bun test` to execute the full test suite
2. If any tests fail, fix the issues before considering the implementation complete
3. Ensure all existing tests still pass and new functionality is properly tested
```

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

### Execute CLI Agent from Custom Path

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Invoke an agent executable located at a non-standard file system path.

```bash
#!/bin/sh
/usr/local/bin/claude --verbose "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
"/c/Program Files/Claude/claude.exe" --verbose "$TRAYCER_PROMPT"
```

```powershell
& "C:\Program Files\Claude\claude.exe" --verbose "$env:TRAYCER_PROMPT"
```

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

### Render Profile Selection UI

Source: https://docs.traycer.ai/extension/tasks/models

Defines styles and renders the profile selection interface, including custom profile handling.

```javascript
const profileButtonStyle = isActive => ({
    padding: "6px 14px",
    fontSize: 14,
    fontWeight: isActive ? 600 : 500,
    cursor: "pointer",
    border: "1px solid rgba(128, 128, 128, 0.2)",
    borderRadius: 4,
    background: isActive ? "rgba(128, 128, 128, 0.1)" : "transparent",
    color: "inherit",
    transition: "all 0.15s ease"
  });
  return <div style={{
    marginTop: 20,
    marginBottom: 24
  }}>
      <div style={{
    marginBottom: 16,
    display: "flex",
    gap: 8,
    alignItems: "center",
    flexWrap: "wrap"
  }}>
        <span style={{
    opacity: 0.7,
    fontSize: 14
  }}>Profile:</span>
        {availableProfiles.map(profile => <button key={profile.id} type="button" style={profileButtonStyle(activeProfileId === profile.id)} onClick={() => updateSelectedSteps({
    ...profile.steps
  }, profile.id)} onMouseEnter={e => activeProfileId !== profile.id && (e.target.style.background = "rgba(128, 128, 128, 0.05)")} onMouseLeave={e => activeProfileId !== profile.id && (e.target.style.background = "transparent")}>
            {formatProfileLabel(profile.name)}
          </button>)}
        {activeProfileId === "custom" ? <span style={{
    padding: "6px 14px",
    fontSize: 14,
    fontWeight: 500,
    border: "1px solid rgba(128, 128, 128, 0.2)",
    borderRadius: 4,
    background: "rgba(128, 128, 128, 0.1)",
    opacity: 0.7
  }}>
            Custom
          </span> : hasCustomProfile() ? <button type="button" style={{
    padding: "6px 14px",
    fontSize: 14,
    fontWeight: 500,
    cursor: "pointer",
    border: "1px solid rgba(128, 128, 128, 0.2)",
    borderRadius: 4,
    background: "transparent",
    color: "inherit",
    opacity: 0.6,
    transition: "all 0.15s ease"
  }} onClick={loadCustomProfile} onMouseEnter={e => {
    e.target.style.background = "rgba(128, 128, 128, 0.05)";
    e.target.style.opacity = 1;
  }} onMouseLeave={e => {
    e.target.style.background = "transparent";
    e.target.style.opacity = 0.6;
  }}>
            Custom
          </button> : null}
      </div>
```

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

### Execute CLI Agent in Verbose Mode

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Run the agent with the --verbose flag to output detailed execution logs.

```bash
#!/bin/sh
aider --yes-always --verbose "$TRAYCER_PROMPT"
```

```powershell
aider --yes-always --verbose "$env:TRAYCER_PROMPT"
```

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

### Calculate Model Costs and Descriptors

Source: https://docs.traycer.ai/extension/tasks/models

Determines the appropriate model descriptor and calculates estimated credits by applying multipliers for token price, reasoning, and caching.

```javascript
const referenceDescriptorId = (costReferenceSteps[feature.id] ?? defaultProfile.steps[feature.id]) ?? modelDescriptors[0]?.id;
    const fallbackDescriptorId = defaultProfile.steps[feature.id] ?? modelDescriptors[0]?.id;
    if (!fallbackDescriptorId) {
      return {
        ...feature,
        descriptor: {
          id: "",
          name: "Unavailable",
          creator: "",
          reasoning: "none",
          input: 1,
          output: 1
        },
        estimatedCredits: feature.baseCredits
      };
    }
    const selectedDescriptorId = selectedSteps[feature.id] ?? fallbackDescriptorId;
    const selectedDescriptor = descriptorById[selectedDescriptorId] ?? descriptorById[fallbackDescriptorId];
    const referenceDescriptor = (descriptorById[referenceDescriptorId] ?? descriptorById[fallbackDescriptorId]) ?? selectedDescriptor;
    if (!selectedDescriptor || !referenceDescriptor) {
      return {
        ...feature,
        descriptor: {
          id: "",
          name: "Unavailable",
          creator: "",
          reasoning: "none",
          input: 1,
          output: 1
        },
        estimatedCredits: feature.baseCredits
      };
    }
    const modelFactor = safeMultiplier(averageTokenPrice(selectedDescriptor) / averageTokenPrice(referenceDescriptor));
    const reasoningFactor = safeMultiplier(getReasoningMultiplier(selectedDescriptor) / getReasoningMultiplier(referenceDescriptor));
    const cacheFactor = safeMultiplier(getCacheModifier(selectedDescriptor) / getCacheModifier(referenceDescriptor));
    const multiplier = modelFactor * reasoningFactor * cacheFactor;
    const modelBaseCost = feature.baseCredits > 0 ? feature.baseCredits - platformAccessFee : 0;
    const estimatedCredits = feature.baseCredits > 0 ? modelBaseCost * multiplier + platformAccessFee : 0;
    return {
      ...feature,
      descriptor: selectedDescriptor,
      estimatedCredits
    };
  });
```

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

### Render Model Selection Table

Source: https://docs.traycer.ai/extension/tasks/models

Displays a table mapping features to model selection dropdowns and their corresponding credit costs.

```jsx
<div>
        <table>
          <thead>
            <tr>
              <th>Feature</th>
              <th>Model</th>
              <th>Credits</th>
            </tr>
          </thead>
          <tbody>
            {rows.map(row => <tr key={row.id}>
                <td>
                  <strong>{row.label}</strong>
                </td>
                <td>
                  <div style={{
    position: "relative"
  }}>
                    <select style={{
    width: "100%",
    padding: "6px 28px 6px 10px",
    fontSize: 14,
    border: "1px solid rgba(128, 128, 128, 0.2)",
    borderRadius: 4,
    background: "transparent",
    color: "inherit",
    cursor: "pointer",
    appearance: "none",
    WebkitAppearance: "none",
    MozAppearance: "none"
  }} name={`model-${row.id}`} aria-label={`${row.label} model`} value={row.descriptor.id} onChange={event => {
    const nextModelId = event.target.value;
    updateSelectedSteps({
      ...selectedSteps,
      [row.id]: nextModelId
    });
  }}>
                      {creators.map(creator => <optgroup key={creator} label={creator.toUpperCase()}>
                          {groupedDescriptors[creator].map(descriptor => <option key={descriptor.id} value={descriptor.id}>
                              {formatModelOption(descriptor)}
                            </option>)}
                        </optgroup>)}
                    </select>
                    <span style={{
    position: "absolute",
    right: 10,
    top: "50%",
    transform: "translateY(-50%)",
    pointerEvents: "none",
    opacity: 0.5
  }}>▼</span>
                  </div>
                </td>
                <td style={{
    whiteSpace: "nowrap"
  }}>
                  <code>{row.estimatedCredits.toFixed(3)}</code>
                </td>
              </tr>)}
          </tbody>
        </table>
      </div>
    </div>;
};
```

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

### Accessing Host Logs via Terminal

Source: https://docs.traycer.ai/reference/troubleshooting

Commands to retrieve or stream host logs directly from the command line.

```bash
traycer host logs --tail 200
```

```bash
traycer host logs --follow
```

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

### Claude Code CLI Basic Commands

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Standard execution commands for the Claude CLI agent across different operating systems.

```bash
#!/bin/sh
claude "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
claude "$TRAYCER_PROMPT"
```

```powershell
claude "$env:TRAYCER_PROMPT"
```

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

### Execute CLI Agent with Auto-commit

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Run the agent with the --auto-commits flag enabled to automatically commit changes.

```bash
#!/bin/sh
aider --yes-always --auto-commits "$TRAYCER_PROMPT"
```

```powershell
aider --yes-always --auto-commits "$env:TRAYCER_PROMPT"
```

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

### Factory Droid CLI Specification Mode

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Execute Factory Droid with unsafe permission skipping enabled.

```bash
#!/bin/sh
droid exec --skip-permissions-unsafe --file "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
droid exec --skip-permissions-unsafe --file "$TRAYCER_PROMPT"
```

```powershell
droid exec --skip-permissions-unsafe --file "$env:TRAYCER_PROMPT"
```

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

### Cursor CLI Force Mode

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Execute cursor-agent with the force flag.

```bash
#!/bin/sh
echo $TRAYCER_PROMPT | cursor-agent --force
```

```bash
#!/bin/sh
echo $TRAYCER_PROMPT | cursor-agent --force
```

```powershell
echo "$env:TRAYCER_PROMPT" | cursor-agent --force
```

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

### Gemini CLI YOLO Mode

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Execute Gemini in YOLO mode using the TRAYCER_PROMPT environment variable.

```bash
#!/bin/sh
gemini --yolo "$TRAYCER_PROMPT"
```

```bash
#!/bin/sh
gemini --yolo "$TRAYCER_PROMPT"
```

```powershell
gemini --yolo "$env:TRAYCER_PROMPT"
```

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

### Aider Auto-approve Usage

Source: https://docs.traycer.ai/extension/integrations/custom-cli-agents

Run Aider with the auto-approve flag enabled.

```bash
#!/bin/sh
aider --yes-always "$TRAYCER_PROMPT"
```
Tokens:
Raw