Lightweight coding agent that runs in your terminal
Rust
Keep Guardian reviewers on summary-based compaction (#43912)
Won Parkdd112a9fd56d
diff --git a/codex-rs/core/src/guardian/reviewer_config.rs b/codex-rs/core/src/guardian/reviewer_config.rsindex a792069ba4..c3624753cc 100644--- a/codex-rs/core/src/guardian/reviewer_config.rs+++ b/codex-rs/core/src/guardian/reviewer_config.rs@@ -13,6 +13,7 @@ use tracing::warn;use crate::config::Config;use crate::config::Constrained;use crate::config::NetworkProxySpec;+use crate::config::TokenBudgetConfig;use super::prompt::BUNDLED_GUARDIAN_POLICY_TEMPLATE;use super::prompt::guardian_policy_prompt_with_config_and_template;@@ -43,6 +44,10 @@ pub fn build_guardian_review_session_config(guardian_config.include_skill_instructions = false;guardian_config.memories.use_memories = false;guardian_config.memories.dedicated_tools = false;+ // Clear inherited startup activation and keep an explicit configuration so model+ // defaults cannot re-enable token-budget mode after the feature is disabled below.+ guardian_config.token_budget_startup_config = None;+ guardian_config.token_budget = Some(TokenBudgetConfig::default());let catalog_auto_review = model_messages.and_then(|messages| messages.auto_review.as_ref());let tenant_policy_config = parent_config.resolve_guardian_policy(model_messages);let policy_template = catalog_auto_review@@ -90,6 +95,8 @@ pub fn build_guardian_review_session_config(Feature::Collab,Feature::MultiAgentV2,Feature::GuardianV2,+ Feature::TokenBudget,+ Feature::ContextManagement,Feature::CodexHooks,Feature::Apps,Feature::Plugins,diff --git a/codex-rs/core/tests/suite/code_mode.rs b/codex-rs/core/tests/suite/code_mode.rsindex 33fe2062de..9e422d0d25 100644--- a/codex-rs/core/tests/suite/code_mode.rs+++ b/codex-rs/core/tests/suite/code_mode.rs@@ -4707,8 +4707,7 @@ await tools.exec_command({ cmd: "true", sandbox_permissions: "require_escalated"#[test_case("node_repl", true, true, false, Some("unbounded"); "text_fallback_without_context_bound")]#[test_case("node_repl", true, true, false, Some("small"); "text_fallback_with_insufficient_context")]#[test_case("node_repl", true, true, false, Some("large_prompt"); "images_resume_after_prompt_pressure")]-#[test_case("node_repl", true, true, false, Some("buffered"); "multimodal_reviewer_preserves_model_owned_fallback_buffer")]-#[test_case("node_repl", true, true, false, Some("rollover"); "multimodal_reviewer_rollover_replays_evidence")]+#[test_case("node_repl", true, true, false, Some("compaction"); "multimodal_reviewer_compaction_preserves_evidence")]#[test_case("cua_repl", false, false, false, None; "cua_disabled")]#[test_case("cua_repl", true, false, false, None; "cua_manually_enabled_text_only")]#[test_case("cua_repl", true, true, false, None; "cua_manually_enabled_multimodal")]@@ -4738,8 +4737,18 @@ async fn code_mode_node_repl_text_evidence_is_visible_only_to_guardian(const PRIVATE_IMAGE: &str = "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR4nGP4z8DwHwAFAAH/iZk9HQAAAABJRU5ErkJggg==";let server = responses::start_mock_server().await;let mcp_server_bin = remote_aware_stdio_server_bin()?;- let reviewer_rollover = reviewer_constraint == Some("rollover");- let reviewer_token_budget = matches!(reviewer_constraint, Some("buffered" | "rollover"));+ let reviewer_compaction = reviewer_constraint == Some("compaction");+ let compact = if reviewer_compaction {+ Some(+ responses::mount_compact_user_history_with_summary_once(+ &server,+ "Guardian browser evidence summary",+ )+ .await,+ )+ } else {+ None+ };let check_detail = enhanced_transcripts && transcript_images && reviewer_constraint.is_none();let mut large_image = Cursor::new(Vec::new());if check_detail {@@ -4762,7 +4771,7 @@ async fn code_mode_node_repl_text_evidence_is_visible_only_to_guardian(.iter_mut().find(|model| model.slug == "gpt-5.6-luna").expect("API-key Guardian reviewer");- if reviewer_token_budget {+ if reviewer_compaction {reviewer.context_window = Some(100_000);reviewer.max_context_window = Some(100_000);reviewer.auto_compact_token_limit = Some(50_000);@@ -4781,7 +4790,8 @@ async fn code_mode_node_repl_text_evidence_is_visible_only_to_guardian(.features.enable(Feature::CodeMode).expect("enable Code Mode");- if reviewer_token_budget {+ if reviewer_compaction {+ config.features.disable(Feature::RemoteCompactionV2).expect("use remote compaction");config.features.enable(Feature::TokenBudget)@@ -4823,8 +4833,7 @@ async fn code_mode_node_repl_text_evidence_is_visible_only_to_guardian(let test = builder.build_with_auto_env(&server).await?;wait_for_mcp_server(&test.codex, repl_server).await?;let images_enabled = auto_review_required || (enhanced_transcripts && transcript_images);- let reviewer_images =- images_enabled && (reviewer_constraint.is_none() || reviewer_token_budget);+ let reviewer_images = images_enabled && (reviewer_constraint.is_none() || reviewer_compaction);let snapshot_padding = if images_enabled && reviewer_constraint != Some("large_prompt") {2_500} else {@@ -4853,13 +4862,13 @@ await tools.mcp__node_repl__js({ code: 'nodeRepl.empty()' });await tools.mcp__node_repl__js({ code: 'await nodeRepl.emitImage(await tab.screenshot())' });if (LARGE_IMAGE) await tools.mcp__node_repl__image_scenario({ scenario: "invalid_image_bytes_then_image" });await tools.exec_command({ cmd: "true", sandbox_permissions: "require_escalated", justification: "review" });-if (!REVIEWER_ROLLOVER) await tools.mcp__node_repl__js({ code: 'await nodeRepl.emitImage(await tab.screenshot())' });+if (!REVIEWER_COMPACTION) await tools.mcp__node_repl__js({ code: 'await nodeRepl.emitImage(await tab.screenshot())' });if (LARGE_IMAGE) await tools.mcp__node_repl__image({});await tools.exec_command({ cmd: "printf second", sandbox_permissions: "require_escalated", justification: "review again" });"#.replace("node_repl", repl_server).replace("SNAPSHOT_PADDING", &snapshot_padding.to_string())- .replace("REVIEWER_ROLLOVER", &reviewer_rollover.to_string())+ .replace("REVIEWER_COMPACTION", &reviewer_compaction.to_string()).replace("LARGE_IMAGE", &check_detail.to_string());let response_mock = responses::mount_sse_sequence(&server,@@ -4888,10 +4897,10 @@ await tools.exec_command({ cmd: "printf second", sandbox_permissions: "require_e]),sse(vec![ev_assistant_message("guardian", r#"{"outcome":"allow"}"#),- if reviewer_token_budget {+ if reviewer_compaction {responses::ev_completed_with_tokens("resp-guardian",- /*total_tokens*/ if reviewer_rollover { 70_000 } else { 60_000 },+ /*total_tokens*/ 70_000,)} else {ev_completed("resp-guardian")@@ -4996,43 +5005,21 @@ await tools.exec_command({ cmd: "printf second", sandbox_permissions: "require_ereviewer_image_urls});- if reviewer_rollover {- let second_request = guardian_requests[1];- let second_prompt = second_request- .message_input_text_groups("user")- .last()- .expect("post-rollover Guardian review prompt")- .join("");- assert!(second_prompt.contains(">>> TRANSCRIPT START\n"));- assert!(!second_prompt.contains(">>> TRANSCRIPT DELTA START\n"));- assert!(second_prompt.contains(NODE_REPL_DOM_MIDDLE));- assert_eq!(- second_request.message_input_image_urls("user"),- vec![PRIVATE_IMAGE.to_string()],- "browser screenshots must be replayed into the fresh reviewer window"- );+ if let Some(compact) = compact {+ let compact_request = compact.single_request();assert!(- second_request- .message_input_texts("developer")- .iter()- .any(|text| text.contains("Previous context window id:")),- "Guardian should roll over before rebuilding browser evidence"+ compact_request+ .message_input_texts("user")+ .concat()+ .contains(NODE_REPL_DOM_MIDDLE));- } else if reviewer_token_budget {- let second_request = guardian_requests[1];- let second_prompt = second_request- .message_input_text_groups("user")- .last()- .expect("buffered Guardian review prompt")- .join("");- assert!(second_prompt.contains(">>> TRANSCRIPT DELTA START\n"));- assert!(- second_request- .message_input_texts("developer")- .iter()- .all(|text| !text.contains("Previous context window id:")),- "Guardian should preserve the reviewer model's fallback buffer before rolling over"+ assert_eq!(+ guardian_requests[1].inputs_of_type("compaction")[0]["encrypted_content"],+ "Guardian browser evidence summary");+ for request in &guardian_requests {+ assert!(!request.has_content_kinds(&["token_budget.context_window"]));+ }}let parent_request = requests.last().expect("parent turn should complete");let parent_input = serde_json::to_string(&parent_request.input())?;diff --git a/codex-rs/core/tests/suite/guardian_review.rs b/codex-rs/core/tests/suite/guardian_review.rsindex 20e8fda4ae..cad3d7ce11 100644--- a/codex-rs/core/tests/suite/guardian_review.rs+++ b/codex-rs/core/tests/suite/guardian_review.rs@@ -32,6 +32,7 @@ use codex_protocol::models::PermissionProfileSnapshot;use codex_protocol::models::ResponseItem;use codex_protocol::openai_models::AutoReviewMessages;use codex_protocol::openai_models::MODEL_SPECIALTY_CYBER;+use codex_protocol::openai_models::ModelTokenBudgetConfig;use codex_protocol::openai_models::ModelsResponse;use codex_protocol::permissions::FileSystemAccessMode;use codex_protocol::permissions::FileSystemPath;@@ -288,10 +289,9 @@ async fn guardian_session_inherits_parent_http_fallback(}#[tokio::test(flavor = "multi_thread", worker_threads = 2)]-#[test_case(60_000, false; "legacy_reviewer_window_is_already_exhausted")]-#[test_case(49_990, true; "retained_followup_reminder_exhausts_reviewer_window")]-async fn guardian_review_resends_full_transcript_after_reviewer_context_rollover(- first_review_total_tokens: i64,+#[test_case(false; "legacy_transcript")]+#[test_case(true; "thread_owned_transcript")]+async fn guardian_review_compacts_with_summary_despite_parent_token_budget(thread_owned: bool,) -> Result<()> {skip_if_no_network!(Ok(()));@@ -301,15 +301,37 @@ async fn guardian_review_resends_full_transcript_after_reviewer_context_rollover);let server = start_mock_server().await;+ let summary = "Guardian retained the user's standing authorization.";+ let compact = core_test_support::responses::mount_compact_user_history_with_summary_once(+ &server, summary,+ )+ .await;let mut builder = test_codex().with_model_info_override("gpt-5.5", |model| {model.auto_review_model_override = Some(model.slug.clone());+ model+ .model_messages+ .as_mut()+ .expect("model messages")+ .token_budget = Some(ModelTokenBudgetConfig {+ enabled: true,+ use_history_notes_extension: true,+ reminder_threshold_tokens: 6_144,+ reminder_message_template: "{n_remaining} tokens remain.".to_string(),+ guidance_message: "Save state before resetting context.".to_string(),+ auto_compact_fallback_prompt: "Save important state.".to_string(),+ auto_compact_fallback_buffer_tokens: 16_384,+ });}).with_config(move |config| {config.features.set_enabled(Feature::GuardianThreadContext, thread_owned).expect("configure Guardian context mode");+ config+ .features+ .disable(Feature::RemoteCompactionV2)+ .expect("use remote compaction");config.model_context_window = Some(100_000);config.model_auto_compact_token_limit = Some(50_000);config.permissions.approval_policy = Constrained::allow_any(AskForApproval::OnRequest);@@ -338,7 +360,7 @@ async fn guardian_review_resends_full_transcript_after_reviewer_context_rolloversse(vec![ev_response_created("resp-guardian-first"),ev_assistant_message("guardian-first", approval),- ev_completed_with_tokens("resp-guardian-first", first_review_total_tokens),+ ev_completed_with_tokens("resp-guardian-first", /*total_tokens*/ 60_000),]),sse(vec![ev_response_created("resp-parent-second"),@@ -373,31 +395,37 @@ async fn guardian_review_resends_full_transcript_after_reviewer_context_rolloverassert_eq!(guardian_requests[0].body_json()["client_metadata"]["thread_id"],guardian_requests[1].body_json()["client_metadata"]["thread_id"],- "the same Guardian reviewer should survive the context-window rollover"+ "the same Guardian reviewer should survive compaction");+ assert!(requests[0].has_content_kinds(&["token_budget.context_window"]));+ for request in &guardian_requests {+ assert!(!request.has_content_kinds(&["token_budget.context_window"]));+ }+ let compact_request = compact.single_request();+ assert!(+ compact_request+ .message_input_texts("user")+ .join("\n")+ .contains(user_authorization)+ );let second_request = guardian_requests[1];- let second_prompt = second_request- .message_input_text_groups("user")- .last()- .expect("post-rollover Guardian review prompt")- .join("");+ assert_eq!(+ second_request.inputs_of_type("compaction")[0]["encrypted_content"],+ summary+ );assert!(second_request- .message_input_texts("developer")- .iter()- .any(|text| text.contains("Previous context window id:")),- "Guardian should have rolled into a new context window"+ .message_input_texts("user")+ .join("\n")+ .contains(user_authorization));- assert!(second_prompt.contains(">>> TRANSCRIPT START\n"));- assert!(!second_prompt.contains(">>> TRANSCRIPT DELTA START\n"));- assert!(second_prompt.contains(user_authorization));assert!(- second_request+ compact_request.message_input_texts("developer").iter().any(|text| text.contains("Use prior reviews as context, not binding precedent.")),- "the follow-up policy reminder should survive reviewer context rollover"+ "the compactor should receive the follow-up policy reminder");Ok(())