Struct MarkdownMetadata 

pub struct MarkdownMetadata {
    pub title: Option<String>,
    pub description: Option<String>,
    pub version: Option<String>,
    pub author: Option<String>,
    pub tags: Vec<String>,
    pub resource_type: Option<String>,
    pub dependencies: Option<BTreeMap<String, Vec<DependencySpec>>>,
    pub extra: BTreeMap<String, Value>,
}

Structured metadata extracted from Markdown frontmatter.

This struct represents all the metadata that can be parsed from YAML or TOML frontmatter in Markdown files. It follows a flexible schema that accommodates both standard AGPM fields and custom extensions.

Standard Fields

The following fields have special meaning in AGPM:

• title: Human-readable name for the resource
• description: Brief explanation of what the resource does
• version: Version identifier (semantic versioning recommended)
• author: Creator or maintainer information
• resource_type: Type classification (“agent” or “snippet”)
• tags: Categorization labels for filtering and discovery
• dependencies: Structured dependencies for transitive resolution

Custom Fields

Additional fields are preserved in the extra map, allowing resource authors to include custom metadata without breaking compatibility.

Serialization

The struct uses Serde for serialization with skip-if-empty optimizations to keep generated frontmatter clean. Empty collections and None values are omitted from the output.

Example

let mut metadata = MarkdownMetadata::default();
metadata.title = Some("Python Linter".to_string());
metadata.version = Some("2.0.1".to_string());
metadata.tags = vec!["python".to_string(), "linting".to_string()];
// Dependencies can be set as a JSON value for the structured format
// This is typically parsed from frontmatter rather than set programmatically

// Custom fields via extra map
let mut extra = BTreeMap::new();
extra.insert("license".to_string(), "MIT".into());
extra.insert("min_python".to_string(), "3.8".into());
metadata.extra = extra;

Fields

title: Option<String>

Human-readable title of the resource.

This is displayed in listings and used for resource identification. If not provided, the title may be extracted from the first heading in the Markdown content.

description: Option<String>

Brief description explaining what the resource does.

Used for documentation and resource discovery. If not provided, the description may be extracted from the first paragraph in the Markdown content.

version: Option<String>

Version identifier for the resource.

Semantic versioning (e.g., “1.2.3”) is recommended for compatibility with dependency resolution, but any string format is accepted.

author: Option<String>

Author or maintainer information.

Can be a name, organization, or contact information. Free-form text.

tags: Vec<String>

Classification tags for categorization and filtering.

Tags help with resource discovery and organization. Common patterns:

• Language-specific: “python”, “javascript”, “rust”
• Functionality: “linting”, “testing”, “documentation”
• Domain: “web-dev”, “data-science”, “devops”

resource_type: Option<String>

Resource type classification.

Currently supported types:

• “agent”: Interactive Claude Code agents
• “snippet”: Code snippets and templates

This field uses rename = "type" to match the frontmatter format while avoiding Rust’s type keyword.

dependencies: Option<BTreeMap<String, Vec<DependencySpec>>>

Dependencies for this resource.

This field uses the structured transitive dependency format where dependencies are organized by resource type (agents, snippets, etc.). Each resource type maps to a list of dependency specifications.

Example:

dependencies:
  agents:
    - path: agents/helper.md
      version: v1.0.0
  snippets:
    - path: snippets/utils.md

extra: BTreeMap<String, Value>

Additional custom metadata fields.

Any frontmatter fields not recognized by the standard schema are preserved here. This allows resource authors to include custom metadata without breaking compatibility with AGPM.

Values are stored as serde_json::Value to handle mixed types (strings, numbers, arrays, objects). Uses BTreeMap for deterministic serialization order.

Implementations

impl MarkdownMetadata

pub fn get_agpm_metadata(&self) -> Option<AgpmMetadata>

Get AGPM-specific metadata from the extra fields.

Extracts the agpm section from the frontmatter if present, which may contain templating flags and nested dependencies.

Trait Implementations

impl Clone for MarkdownMetadata

fn clone(&self) -> MarkdownMetadata

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MarkdownMetadata

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for MarkdownMetadata

fn default() -> MarkdownMetadata

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for MarkdownMetadata

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for MarkdownMetadata

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.