feat(review): ingest linked document context

Author: haasonsaas

TODO.md

@@ -79,7 +79,7 @@ This roadmap is derived from deep research into Greptile's public docs, blog, MC
 43. [x] Add structured UI editing for custom context notes, files, and scopes.
 44. [x] Add per-path scoped review instructions in the Settings UI for common repo areas.
 45. [x] Support Jira/Linear issue context ingestion for PR-linked reviews.
-46. [ ] Support document-backed context ingestion for design docs, RFCs, and runbooks.
+46. [x] Support document-backed context ingestion for design docs, RFCs, and runbooks.
 47. [x] Add explicit "intent mismatch" review checks comparing PR changes to ticket acceptance criteria.
 48. [x] Add review artifacts that show which external context sources influenced a finding.
 49. [x] Add tests for pattern repository resolution across local paths, Git URLs, and broken sources.

src/config.rs

@@ -155,6 +155,15 @@ pub struct LinkedIssueContext {
     pub summary: String,
 }
 
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
+pub struct DocumentContext {
+    pub source: String,
+    pub title: String,
+    pub url: String,
+    #[serde(default)]
+    pub summary: String,
+}
+
 #[derive(Debug, Clone, Serialize, Deserialize, Default)]
 pub struct AutomationConfig {
     /// Outbound webhook URL for downstream automation consumers.
@@ -501,6 +510,9 @@ pub struct Config {
     #[serde(default, skip_serializing_if = "Vec::is_empty")]
     pub linked_issue_contexts: Vec<LinkedIssueContext>,
 
+    #[serde(default, skip_serializing_if = "Vec::is_empty")]
+    pub document_contexts: Vec<DocumentContext>,
+
     #[serde(default)]
     pub pattern_repositories: Vec<PatternRepositoryConfig>,
 
@@ -728,6 +740,7 @@ impl Default for Config {
             paths: HashMap::new(),
             custom_context: Vec::new(),
             linked_issue_contexts: Vec::new(),
+            document_contexts: Vec::new(),
             pattern_repositories: Vec::new(),
             rules_files: Vec::new(),
             max_active_rules: default_max_active_rules(),

src/core/context_provenance.rs

@@ -10,6 +10,10 @@ pub enum ContextProvenance {
         name: String,
     },
     CustomContextNotes,
+    DocumentContext {
+        source: String,
+        title: String,
+    },
     JiraIssueContext {
         issue_key: String,
     },
@@ -65,6 +69,13 @@ impl ContextProvenance {
         }
     }
 
+    pub fn document_context(source: impl Into<String>, title: impl Into<String>) -> Self {
+        Self::DocumentContext {
+            source: source.into(),
+            title: title.into(),
+        }
+    }
+
     pub fn linear_issue_context(issue_id: impl Into<String>) -> Self {
         Self::LinearIssueContext {
             issue_id: issue_id.into(),
@@ -118,6 +129,7 @@ impl ContextProvenance {
             }
             Self::Analyzer { .. }
             | Self::CustomContextNotes
+            | Self::DocumentContext { .. }
             | Self::JiraIssueContext { .. }
             | Self::LinearIssueContext { .. }
             | Self::DependencyGraphNeighborhood
@@ -141,6 +153,7 @@ impl ContextProvenance {
         let source = match self {
             Self::ActiveReviewRules | Self::Analyzer { .. } => return None,
             Self::CustomContextNotes => "custom-context".to_string(),
+            Self::DocumentContext { source, .. } => source.trim().to_string(),
             Self::JiraIssueContext { .. } => "jira-issue".to_string(),
             Self::LinearIssueContext { .. } => "linear-issue".to_string(),
             Self::DependencyGraphNeighborhood => "dependency-graph".to_string(),
@@ -165,6 +178,9 @@ impl ContextProvenance {
             Self::ActiveReviewRules => "active review rules".to_string(),
             Self::Analyzer { name } => format!("{name} analyzer"),
             Self::CustomContextNotes => "custom context notes".to_string(),
+            Self::DocumentContext { source, title } => {
+                format!("document context ({source}): {title}")
+            }
             Self::JiraIssueContext { issue_key } => format!("jira issue context: {issue_key}"),
             Self::LinearIssueContext { issue_id } => {
                 format!("linear issue context: {issue_id}")
@@ -260,6 +276,10 @@ mod tests {
             ContextProvenance::RepositoryGraphMetadata.to_string(),
             "repository graph metadata"
         );
+        assert_eq!(
+            ContextProvenance::document_context("design-doc", "Checkout architecture").to_string(),
+            "document context (design-doc): Checkout architecture"
+        );
         assert_eq!(
             ContextProvenance::jira_issue_context("ENG-123").to_string(),
             "jira issue context: ENG-123"
@@ -290,6 +310,12 @@ mod tests {
                 .as_deref(),
             Some("context-source:symbol-graph")
         );
+        assert_eq!(
+            ContextProvenance::document_context("runbook", "Pager escalation")
+                .artifact_tag()
+                .as_deref(),
+            Some("context-source:runbook")
+        );
         assert_eq!(
             ContextProvenance::jira_issue_context("ENG-123")
                 .artifact_tag()

src/review/context_helpers.rs

@@ -6,7 +6,8 @@ mod pattern_repositories;
 mod ranking;
 
 pub use injection::{
-    inject_custom_context, inject_linked_issue_context, inject_pattern_repository_context,
+    inject_custom_context, inject_document_context, inject_linked_issue_context,
+    inject_pattern_repository_context,
 };
 pub use pattern_repositories::{resolve_pattern_repositories, PatternRepositoryMap};
 pub use ranking::rank_and_trim_context_chunks;

src/review/context_helpers/injection.rs

@@ -92,6 +92,45 @@ pub fn inject_linked_issue_context(
     }
 }
 
+fn document_source_label(source: &str) -> &str {
+    match source {
+        "design-doc" => "design doc",
+        "rfc" => "RFC",
+        "runbook" => "runbook",
+        _ => "document",
+    }
+}
+
+pub fn inject_document_context(
+    config: &config::Config,
+    diff: &core::UnifiedDiff,
+    context_chunks: &mut Vec<core::LLMContextChunk>,
+) {
+    for document in &config.document_contexts {
+        let mut lines = vec![format!(
+            "Linked {} context: {}",
+            document_source_label(&document.source),
+            document.title
+        )];
+
+        if !document.url.trim().is_empty() {
+            lines.push(format!("URL: {}", document.url.trim()));
+        }
+        if !document.summary.trim().is_empty() {
+            lines.push("Relevant design / runbook context:".to_string());
+            lines.push(document.summary.trim().to_string());
+        }
+
+        context_chunks.push(
+            core::LLMContextChunk::documentation(diff.file_path.clone(), lines.join("\n"))
+                .with_provenance(core::ContextProvenance::document_context(
+                    document.source.clone(),
+                    document.title.clone(),
+                )),
+        );
+    }
+}
+
 pub async fn inject_pattern_repository_context(
     config: &config::Config,
     resolved_repositories: &PatternRepositoryMap,
@@ -251,6 +290,32 @@ mod tests {
             .contains("Deployment manifests should use the new secret name."));
     }
 
+    #[test]
+    fn inject_document_context_uses_document_provenance() {
+        let mut config = config::Config::default();
+        config.document_contexts = vec![config::DocumentContext {
+            source: "design-doc".to_string(),
+            title: "Checkout resiliency RFC".to_string(),
+            url: "https://github.com/evalops/diffscope/blob/main/docs/rfcs/checkout.md".to_string(),
+            summary: "Keep retries idempotent and preserve queue ordering.".to_string(),
+        }];
+
+        let mut context_chunks = Vec::new();
+        inject_document_context(&config, &diff_for("src/lib.rs"), &mut context_chunks);
+
+        assert_eq!(context_chunks.len(), 1);
+        assert!(context_chunks[0]
+            .content
+            .contains("Linked design doc context: Checkout resiliency RFC"));
+        assert_eq!(
+            context_chunks[0].provenance,
+            Some(core::ContextProvenance::document_context(
+                "design-doc",
+                "Checkout resiliency RFC"
+            ))
+        );
+    }
+
     #[tokio::test]
     async fn inject_pattern_repository_context_tags_source_and_chunks() {
         let dir = tempfile::tempdir().unwrap();

src/review/pipeline/file_context/sources/repo.rs

@@ -14,6 +14,11 @@ pub(in super::super) async fn inject_repository_context(
         diff,
         context_chunks,
     );
+    super::super::super::super::context_helpers::inject_document_context(
+        &services.config,
+        diff,
+        context_chunks,
+    );
     super::super::super::super::context_helpers::inject_custom_context(
         &services.config,
         &services.context_fetcher,