feat(core): OSCAL 1.1 catalog types + mapping to controls corpus

Deserialise the OSCAL catalog served by breakpilot-compliance and flatten it
into the framework-agnostic traits::Control the mapping engine consumes. Pure
serde (no reqwest); exposes content-hash for snapshot/drift on the consumer side.
Fixture-tested against the real 40-control CRA catalog (3 lib tests).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Sharang Parnerkar
2026-07-20 17:56:39 +02:00
co-authored by Claude Opus 4.8
parent 94c0d51a11
commit 0e5a2d7e43
3 changed files with 3017 additions and 0 deletions
+2
View File
@@ -10,6 +10,7 @@ pub mod mcp;
pub mod mcp_token;
pub mod notification;
pub mod onboarding;
pub mod oscal;
pub mod pentest;
pub mod repository;
pub mod sbom;
@@ -39,6 +40,7 @@ pub use onboarding::{
GitArtifactConfig, IssueTrackerConfig, OnboardedTarget, PlcArtifactConfig, PlcFormat,
TargetScanConfig, TargetType, TargetTypeCandidate, WebArtifactConfig,
};
pub use oscal::OscalDocument;
pub use pentest::{
AttackChainNode, AttackNodeStatus, AuthMode, CodeContextHint, Environment, IdentityProvider,
PentestAuthConfig, PentestConfig, PentestEvent, PentestMessage, PentestSession, PentestStats,
+250
View File
@@ -0,0 +1,250 @@
//! OSCAL 1.1 catalog types + mapping into the controls corpus.
//!
//! Deserialises the OSCAL catalog served by breakpilot-compliance
//! (`GET /api/compliance/v1/oscal/catalog`) and maps its controls into the
//! framework-agnostic [`crate::traits::Control`] that the mapping engine consumes.
//! Only the fields we use are modelled; unknown OSCAL fields are ignored so the
//! producer can add detail without breaking us.
//!
//! Scope boundary: this is the *catalog* (domain content). Assessment objectives
//! and scanner routing live in our assessment layer, not here — see
//! [`crate::traits::ControlsProvider`].
use serde::Deserialize;
use crate::models::onboarding::ComplianceFramework;
use crate::traits::Control as CorpusControl;
/// A parsed OSCAL catalog document (`{"catalog": {...}}`).
#[derive(Debug, Clone, Deserialize)]
pub struct OscalDocument {
pub catalog: Catalog,
}
/// An OSCAL catalog: metadata + a tree of control groups.
#[derive(Debug, Clone, Deserialize)]
pub struct Catalog {
pub uuid: String,
pub metadata: Metadata,
#[serde(default)]
pub groups: Vec<Group>,
#[serde(rename = "back-matter", default)]
pub back_matter: Option<BackMatter>,
}
/// Catalog metadata (title/version + provenance props).
#[derive(Debug, Clone, Deserialize)]
pub struct Metadata {
pub title: String,
pub version: String,
#[serde(rename = "oscal-version")]
pub oscal_version: String,
#[serde(default)]
pub props: Vec<Prop>,
}
/// A name/value property, optionally namespaced.
#[derive(Debug, Clone, Deserialize)]
pub struct Prop {
pub name: String,
pub value: String,
#[serde(default)]
pub ns: Option<String>,
}
/// A control group (may nest sub-groups and controls).
#[derive(Debug, Clone, Deserialize)]
pub struct Group {
#[serde(default)]
pub id: String,
#[serde(default)]
pub title: String,
#[serde(default)]
pub controls: Vec<Control>,
#[serde(default)]
pub groups: Vec<Group>,
}
/// An OSCAL control (may nest enhancement controls).
#[derive(Debug, Clone, Deserialize)]
pub struct Control {
pub id: String,
#[serde(default)]
pub title: String,
#[serde(default)]
pub props: Vec<Prop>,
#[serde(default)]
pub parts: Vec<Part>,
#[serde(default)]
pub links: Vec<Link>,
#[serde(default)]
pub controls: Vec<Control>,
}
/// A control part (e.g. the `statement`), may nest sub-parts.
#[derive(Debug, Clone, Deserialize)]
pub struct Part {
#[serde(default)]
pub name: String,
#[serde(default)]
pub prose: Option<String>,
#[serde(default)]
pub parts: Vec<Part>,
}
/// A link, e.g. a `reference` to a back-matter resource.
#[derive(Debug, Clone, Deserialize)]
pub struct Link {
pub href: String,
#[serde(default)]
pub rel: Option<String>,
}
/// Back-matter holding referenced resources (e.g. the CRA measures).
#[derive(Debug, Clone, Deserialize)]
pub struct BackMatter {
#[serde(default)]
pub resources: Vec<Resource>,
}
/// A back-matter resource referenced by control links.
#[derive(Debug, Clone, Deserialize)]
pub struct Resource {
pub uuid: String,
#[serde(default)]
pub title: Option<String>,
#[serde(default)]
pub description: Option<String>,
}
impl Metadata {
/// First prop value with the given name.
pub fn prop(&self, name: &str) -> Option<&str> {
self.props
.iter()
.find(|p| p.name == name)
.map(|p| p.value.as_str())
}
}
impl Control {
/// First prop value with the given name.
pub fn prop(&self, name: &str) -> Option<&str> {
self.props
.iter()
.find(|p| p.name == name)
.map(|p| p.value.as_str())
}
/// The control's `statement` prose, if present.
pub fn statement(&self) -> Option<&str> {
self.parts
.iter()
.find(|p| p.name == "statement")
.and_then(|p| p.prose.as_deref())
}
}
impl OscalDocument {
/// The framework this catalog declares (`metadata.props[name="framework"]`).
pub fn framework(&self) -> Option<ComplianceFramework> {
framework_from_str(self.catalog.metadata.prop("framework")?)
}
/// The catalog `content-hash` prop — consumers pin this to snapshot/detect drift.
pub fn content_hash(&self) -> Option<&str> {
self.catalog.metadata.prop("content-hash")
}
/// Flatten the catalog into the corpus controls the mapping engine consumes.
pub fn to_controls(&self) -> Vec<CorpusControl> {
let framework = self.framework().unwrap_or(ComplianceFramework::Cra);
let source_label = self.catalog.metadata.title.as_str();
let mut out = Vec::new();
for group in &self.catalog.groups {
collect_group(group, framework, source_label, &mut out);
}
out
}
}
/// Map an OSCAL framework token (e.g. `"cra"`) to [`ComplianceFramework`] via its
/// serde snake_case representation.
fn framework_from_str(raw: &str) -> Option<ComplianceFramework> {
serde_json::from_value(serde_json::Value::String(raw.to_string())).ok()
}
fn collect_group(
group: &Group,
framework: ComplianceFramework,
source_label: &str,
out: &mut Vec<CorpusControl>,
) {
for control in &group.controls {
collect_control(control, framework, source_label, out);
}
for sub in &group.groups {
collect_group(sub, framework, source_label, out);
}
}
fn collect_control(
control: &Control,
framework: ComplianceFramework,
source_label: &str,
out: &mut Vec<CorpusControl>,
) {
let source = match control.prop("annex-anchor") {
Some(anchor) => Some(format!("{source_label} · {anchor}")),
None => Some(source_label.to_string()),
};
out.push(CorpusControl {
id: control.id.clone(),
framework,
title: control.title.clone(),
text: control.statement().unwrap_or_default().to_string(),
source,
});
for enhancement in &control.controls {
collect_control(enhancement, framework, source_label, out);
}
}
#[cfg(test)]
#[allow(clippy::unwrap_used)]
mod tests {
use super::*;
const CATALOG: &str = include_str!("../../tests/data/cra_catalog.json");
fn parse() -> OscalDocument {
serde_json::from_str(CATALOG).unwrap()
}
#[test]
fn parses_full_catalog() {
let doc = parse();
assert_eq!(doc.catalog.metadata.oscal_version, "1.1.2");
assert!(!doc.catalog.groups.is_empty());
assert!(doc.catalog.back_matter.is_some());
}
#[test]
fn maps_all_controls_to_corpus() {
let doc = parse();
let controls = doc.to_controls();
assert_eq!(controls.len(), 40);
assert_eq!(doc.framework(), Some(ComplianceFramework::Cra));
let c8 = controls.iter().find(|c| c.id == "cra-ai-8").unwrap();
assert_eq!(c8.framework, ComplianceFramework::Cra);
assert!(!c8.title.is_empty());
assert!(!c8.text.is_empty(), "statement prose should map into text");
assert!(c8.source.as_deref().unwrap_or_default().contains("Annex I"));
}
#[test]
fn exposes_content_hash_for_snapshotting() {
assert_eq!(parse().content_hash().map(str::len), Some(64));
}
}
File diff suppressed because it is too large Load Diff