Module tools 

Agent tools using Rig’s Tool trait

These tools wrap existing CLI functionality for the agent to use.

Available Tools

File Operations

• ReadFileTool - Read file contents
• WriteFileTool - Write single files (Dockerfiles, Terraform, etc.)
• WriteFilesTool - Write multiple files (Terraform modules, Helm charts)
• ListDirectoryTool - List directory contents

Analysis

• AnalyzeTool - Analyze project architecture, dependencies, build commands

Security

• SecurityScanTool - Security vulnerability scanning
• VulnerabilitiesTool - Dependency vulnerability checking

Linting

• HadolintTool - Native Dockerfile linting (best practices, security)
• DclintTool - Native Docker Compose linting (best practices, style, security)
• HelmlintTool - Native Helm chart structure/template linting
• KubelintTool - Native Kubernetes manifest security/best practice linting

Resource Optimization

• K8sOptimizeTool - Kubernetes resource right-sizing and cost optimization
• K8sCostsTool - Kubernetes workload cost attribution and analysis
• K8sDriftTool - Detect configuration drift between manifests and cluster

Prometheus Integration (for live K8s analysis)

• PrometheusDiscoverTool - Discover Prometheus services in Kubernetes cluster
• PrometheusConnectTool - Establish connection to Prometheus (port-forward or URL)
• BackgroundProcessManager - Manage long-running background processes

Helm vs Kubernetes Linting

• HelmlintTool: Use for Helm chart development - validates Chart.yaml, values.yaml, Go template syntax, and Helm-specific best practices. Works on chart directories.
• KubelintTool: Use for K8s security - checks rendered manifests for privileged containers, missing probes, RBAC issues, resource limits. Works on YAML files, Helm charts (renders them), and Kustomize directories.

Diagnostics

• DiagnosticsTool - Check for code errors via IDE/LSP or language-specific commands

Terraform

• TerraformFmtTool - Format Terraform configuration files
• TerraformValidateTool - Validate Terraform configurations
• TerraformInstallTool - Install Terraform CLI (auto-detects OS)

Shell

• ShellTool - Execute validation commands (docker build, terraform validate, helm lint)

Planning (Forge-style workflow)

• PlanCreateTool - Create structured plan files with task checkboxes
• PlanNextTool - Get next pending task and mark it in-progress
• PlanUpdateTool - Update task status (done, failed)
• PlanListTool - List all available plan files

Web

• WebFetchTool - Fetch content from URLs (converts HTML to markdown)

Platform (Syncable Platform API)

• ListOrganizationsTool - List organizations the user belongs to
• ListProjectsTool - List projects within an organization
• SelectProjectTool - Select a project as current context
• CurrentContextTool - Get the currently selected project context
• OpenProviderSettingsTool - Open cloud provider settings in browser
• CheckProviderConnectionTool - Check if a cloud provider is connected
• ListDeploymentConfigsTool - List deployment configurations for a project
• TriggerDeploymentTool - Trigger a deployment using a config
• GetDeploymentStatusTool - Get deployment task status and progress
• ListDeploymentsTool - List recent deployments with URLs
• GetServiceLogsTool - Get container logs for a deployed service

Error Handling Pattern

Tools use the shared error utilities in error.rs:

1. Each tool keeps its own error type (e.g., ReadFileError, ShellError)
2. Use ToolErrorContext trait to add context when propagating errors
3. Use format_error_for_llm for structured JSON error responses to the agent
4. Error categories help the agent understand and recover from errors

See error.rs for the complete error handling infrastructure.

Response Format Pattern

Tools use the shared response utilities in response.rs for consistent output:

1. Use format_file_content for file read operations (with truncation metadata)
2. Use format_list for directory listings and search results
3. Use format_write_success for successful write operations
4. Use format_cancelled for user-cancelled operations
5. Use ResponseMetadata to track truncation, compression, and item counts

For large outputs (analysis, lint results), use compress_tool_output or compress_analysis_output from compression.rs which store full data and return a compressed summary with retrieval reference.

Example

use crate::agent::tools::response::{format_file_content, format_list};

// File read response
Ok(format_file_content(&path, &content, total_lines, returned_lines, truncated))

// Directory listing response
Ok(format_list(&path, &entries, total_count, was_truncated))

See response.rs for the complete response formatting infrastructure.

Re-exports

pub use compression::CompressionConfig;

pub use compression::compress_analysis_output;

pub use compression::compress_tool_output;

pub use error::ErrorCategory;

pub use error::ToolErrorContext;

pub use error::detect_error_category;

pub use error::format_error_for_llm;

pub use error::format_error_with_context;

pub use response::ResponseMetadata;

pub use response::ToolResponse;

pub use response::format_cancelled;

pub use response::format_file_content;

pub use response::format_file_content_range;

pub use response::format_list;

pub use response::format_list_with_metadata;

pub use response::format_success;

pub use response::format_success_with_metadata;

pub use response::format_write_success;

pub use background::BackgroundProcessManager;

pub use platform::CheckProviderConnectionTool;

pub use platform::CreateDeploymentConfigTool;

pub use platform::CurrentContextTool;

pub use platform::DeployServiceTool;

pub use platform::GetDeploymentStatusTool;

pub use platform::GetServiceLogsTool;

pub use platform::ListDeploymentCapabilitiesTool;

pub use platform::ListDeploymentConfigsTool;

pub use platform::ListDeploymentsTool;

pub use platform::ListOrganizationsTool;

pub use platform::ListProjectsTool;

pub use platform::OpenProviderSettingsTool;

pub use platform::SelectProjectTool;

pub use platform::TriggerDeploymentTool;

Modules

background

Background Process Manager

compression

Smart Context Compression for Tool Outputs

error

Common error utilities for agent tools

output_store

RAG Storage Layer for Tool Outputs

platform

Platform tools for managing Syncable platform resources

response

Response formatting utilities for agent tools

Structs

AnalyzeTool

Tool to analyze a project

DclintTool

Tool to lint Docker Compose files natively

DiagnosticsTool

HadolintTool

Tool to lint Dockerfiles natively

HelmlintTool

Tool to lint Helm charts natively

K8sCostsTool

Tool for analyzing Kubernetes workload costs

K8sDriftTool

Tool for detecting Kubernetes configuration drift

K8sOptimizeTool

Tool for analyzing Kubernetes resource configurations

KubelintTool

Tool to lint Kubernetes manifests natively

ListDirectoryTool

ListOutputsTool

Tool to list all available stored outputs

PlanCreateTool

PlanListTool

PlanNextTool

PlanUpdateTool

PrometheusConnectTool

Tool for connecting to Prometheus

PrometheusDiscoverTool

Tool for discovering Prometheus in Kubernetes clusters

ReadFileTool

RetrieveOutputTool

Tool to retrieve detailed data from compressed tool outputs

SecurityScanTool

ShellTool

TerraformFmtTool

Tool to format Terraform configurations

TerraformInstallTool

Tool to install Terraform CLI

TerraformValidateTool

Tool to validate Terraform configurations

TruncationLimits

Configuration for output truncation limits

VulnerabilitiesTool

WebFetchTool

WriteFileTool

WriteFilesTool

Functions

truncate_json_output

Truncate JSON output to fit within context limits. Intelligently summarizes large arrays and nested objects.