Expand Rust E2E coverage

[stephentoub]

The Rust SDK had much less replay-backed E2E coverage than the other SDKs, which made protocol and lifecycle regressions harder to catch. This brings Rust up to parity with the current .NET E2E suite and ports the latest upstream empty-session fork behavior.

Summary

• Adds a single serialized Rust E2E integration binary with a shared replay-proxy harness and one Rust test case per current .NET E2E test.
• Adds the required replay snapshots and folds the older mode-handler E2E coverage into the new suite.
• Updates Rust SDK internals needed for parity, including SessionFs routing during create/resume, session ID consistency checks, JSON-RPC force-close behavior, serialized model-list cache fills, and cancellation-safe pending session registration cleanup.

Notes

• The Rust E2E suite is intentionally run as one serialized integration binary to avoid parallel replay-proxy and snapshot races.
• The only ignored Rust E2Es match the two upstream-skipped .NET cases.
• Empty-session fork handling accepts both the older no-persisted-events error and the newer successful empty fork behavior.

Validation

• cargo +nightly-2026-04-14 fmt --all -- --config-path .rustfmt.nightly.toml --check
• cargo clippy --all-targets --features test-support -- --no-deps -D warnings -D clippy::unwrap_used -D clippy::disallowed_macros -D clippy::await_holding_invalid_type
• cargo test -q --features test-support --lib
• cargo test -q --features test-support --test e2e --no-run
• cargo test --features test-support --test e2e -- --test-threads=1 (342 passed; 0 failed; 2 ignored)
• Focused post-review regression tests for list_models, pending session cancellation cleanup, and the auto-mode-switch E2E

[Copilot]

Pull request overview

Expands the Rust SDK’s replay-backed E2E suite to better match coverage in other SDKs, and updates Rust internals around session lifecycle, JSON-RPC shutdown semantics, and model listing to support the new tests and reduce regression risk.

Changes:

• Added a large set of Rust replay-backed E2E tests (plus new replay snapshots) spanning session lifecycle, RPC surfaces, tools/hooks, telemetry, and multi-client behavior.
• Updated Rust session create/resume internals to pre-register sessions, enforce session-id consistency, and ensure cancellation-safe cleanup.
• Improved client runtime behavior via JSON-RPC force-close support and a single-flight cached list_models implementation; added reasoning_effort to ResumeSessionConfig.

Show a summary per file

File	Description
test/snapshots/session_config/should_use_custom_session_id.yaml	Adds replay snapshot for session-config E2E coverage.
test/snapshots/session_config/should_create_session_with_custom_provider_config.yaml	Adds replay snapshot for session-config E2E coverage.
test/snapshots/session_config/should_apply_reasoning_effort_on_session_create.yaml	Adds replay snapshot for session-config E2E coverage.
test/snapshots/session_config/should_apply_all_reasoning_effort_values_on_session_create.yaml	Adds replay snapshot for session-config E2E coverage.
test/snapshots/rust_multi_client/two_clients_register_different_tools_and_agent_uses_both.yaml	Adds replay snapshot for Rust multi-client tool registration behavior.
test/snapshots/rust_multi_client/disconnecting_client_removes_its_tools.yaml	Adds replay snapshot for multi-client tool removal behavior.
test/snapshots/rust_multi_client/both_clients_see_tool_request_and_completion_events.yaml	Adds replay snapshot for multi-client tool event fan-out.
test/snapshots/rpc_tasks_and_handlers/should_return_expected_results_for_missing_pending_handler_requestids.yaml	Adds replay snapshot for tasks/handlers RPC E2E coverage.
test/snapshots/rpc_tasks_and_handlers/should_report_implemented_error_for_missing_task_agent_type.yaml	Adds replay snapshot for tasks/handlers RPC E2E coverage.
test/snapshots/rpc_tasks_and_handlers/should_report_implemented_error_for_invalid_task_agent_model.yaml	Adds replay snapshot for tasks/handlers RPC E2E coverage.
test/snapshots/rpc_tasks_and_handlers/should_list_task_state_and_return_false_for_missing_task_operations.yaml	Adds replay snapshot for tasks/handlers RPC E2E coverage.
test/snapshots/rpc_shell_edge_cases/shell_kill_unknown_processid_returns_false.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_kill_cleans_up_after_terminating_signal.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_exec_with_timeout_kills_long_running_command.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_exec_with_stderr_output_cleans_up.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_exec_with_nonexistent_command_returns_processid_and_cleans_up.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_exec_with_large_stdout_cleans_up.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_edge_cases/shell_exec_with_custom_cwd_honors_override.yaml	Adds replay snapshot for shell edge-case RPC coverage.
test/snapshots/rpc_shell_and_fleet/should_kill_shell_process.yaml	Adds replay snapshot for shell/fleet RPC coverage.
test/snapshots/rpc_shell_and_fleet/should_execute_shell_command.yaml	Adds replay snapshot for shell/fleet RPC coverage.
test/snapshots/rpc_session_state/should_update_existing_workspace_file_with_update_operation.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_set_and_get_each_session_mode_value.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_report_implemented_errors_for_unsupported_session_rpc_paths.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_report_error_reading_nonexistent_workspace_file.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_reject_workspace_file_path_traversal.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_reject_empty_or_whitespace_session_name.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_read_update_and_delete_plan.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_handle_forking_session_without_persisted_events.yaml	Adds replay snapshot for session-state fork behavior coverage.
test/snapshots/rpc_session_state/should_get_and_set_session_mode.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_get_and_set_session_metadata.yaml	Adds replay snapshot for session-state RPC coverage.
test/snapshots/rpc_session_state/should_emit_title_changed_event_each_time_name_set_is_called.yaml	Adds replay snapshot for session-state event coverage.
test/snapshots/rpc_session_state/should_create_workspace_file_with_nested_path_auto_creating_dirs.yaml	Adds replay snapshot for workspace file RPC coverage.
test/snapshots/rpc_session_state/should_call_workspace_file_rpc_methods.yaml	Adds replay snapshot for workspace file RPC coverage.
test/snapshots/rpc_session_state/should_call_session_usage_and_permission_rpcs.yaml	Adds replay snapshot for session usage/permission RPC coverage.
test/snapshots/rpc_session_state/should_call_session_rpc_model_switchto.yaml	Adds replay snapshot for model RPC coverage.
test/snapshots/rpc_session_state/should_call_session_rpc_model_getcurrent.yaml	Adds replay snapshot for model RPC coverage.
test/snapshots/rpc_server/should_discover_server_mcp_and_skills.yaml	Adds replay snapshot for server-scoped RPC coverage.
test/snapshots/rpc_server/should_call_rpc_tools_list_with_typed_result.yaml	Adds replay snapshot for server-scoped RPC coverage.
test/snapshots/rpc_server/should_call_rpc_ping_with_typed_params_and_result.yaml	Adds replay snapshot for server-scoped RPC coverage.
test/snapshots/rpc_server/should_call_rpc_models_list_with_typed_result.yaml	Adds replay snapshot for server-scoped RPC coverage.
test/snapshots/rpc_server/should_call_rpc_account_get_quota_when_authenticated.yaml	Adds replay snapshot for server-scoped RPC coverage.
test/snapshots/rpc_mcp_config/should_round_trip_http_mcp_oauth_config_rpc.yaml	Adds replay snapshot for MCP config RPC coverage.
test/snapshots/rpc_mcp_config/should_call_server_mcp_config_rpcs.yaml	Adds replay snapshot for MCP config RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_report_error_when_mcp_oauth_server_is_not_remote.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_report_error_when_mcp_oauth_server_is_not_configured.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_report_error_when_mcp_host_is_not_initialized.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_report_error_when_extensions_are_not_available.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_reload_session_skills.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_list_plugins.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_list_mcp_servers_with_configured_server.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_list_extensions.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_mcp_and_skills/should_list_and_toggle_session_skills.yaml	Adds replay snapshot for MCP/skills RPC coverage.
test/snapshots/rpc_event_side_effects/should_emit_workspace_file_changed_event_when_file_created.yaml	Adds replay snapshot for event side-effects coverage.
test/snapshots/rpc_event_side_effects/should_emit_title_changed_event_when_name_set.yaml	Adds replay snapshot for event side-effects coverage.
test/snapshots/rpc_event_side_effects/should_emit_plan_changed_update_operation_on_second_update.yaml	Adds replay snapshot for event side-effects coverage.
test/snapshots/rpc_event_side_effects/should_emit_plan_changed_event_for_update_and_delete.yaml	Adds replay snapshot for event side-effects coverage.
test/snapshots/rpc_event_side_effects/should_emit_mode_changed_event_when_mode_set.yaml	Adds replay snapshot for event side-effects coverage.
test/snapshots/rpc_agents/should_select_and_get_current_agent.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_agents/should_return_null_when_no_agent_is_selected.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_agents/should_return_empty_list_when_no_custom_agents_configured.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_agents/should_list_available_custom_agents.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_agents/should_emit_subagent_selected_and_deselected_events.yaml	Adds replay snapshot for agent event coverage.
test/snapshots/rpc_agents/should_deselect_current_agent.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_agents/should_call_agent_reload.yaml	Adds replay snapshot for agent RPC coverage.
test/snapshots/rpc_additional_edge_cases/workspaces_getworkspace_returns_stable_result_across_calls.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/workspaces_createfile_then_listfiles_returns_sorted_or_stable_order.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/workspaces_create_file_with_unicode_content_round_trips.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/workspaces_create_file_with_large_content_round_trips.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/workspaces_create_file_with_empty_content_round_trips.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/usage_get_metrics_on_fresh_session_returns_zero_tokens.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/shell_exec_with_zero_timeout_does_not_kill_long_running_command.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/plan_update_with_empty_content_then_read_returns_empty.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/plan_delete_when_none_exists_is_idempotent.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/permissions_set_approve_all_toggle_round_trips.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/permissions_reset_session_approvals_on_fresh_session_is_noop.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/name_set_with_unicode_round_trips.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/rpc_additional_edge_cases/mode_set_to_same_value_multiple_times_stays_stable.yaml	Adds replay snapshot for additional edge-case RPC coverage.
test/snapshots/per-session-auth/session_uses_client_token_when_no_session_token_is_supplied.yaml	Adds replay snapshot for per-session auth E2E coverage.
test/snapshots/per-session-auth/session_token_overrides_client_token.yaml	Adds replay snapshot for per-session auth E2E coverage.
test/snapshots/per-session-auth/session_fails_with_invalid_token.yaml	Adds replay snapshot for per-session auth E2E coverage.
test/snapshots/per-session-auth/session_auth_status_is_unauthenticated_without_token.yaml	Adds replay snapshot for per-session auth E2E coverage.
test/snapshots/multi_client_commands_elicitation/client_receives_commands_changed_when_another_client_joins_with_commands.yaml	Adds replay snapshot for multi-client commands/elicitation coverage.
test/snapshots/multi_client_commands_elicitation/capabilities_changed_fires_when_second_client_joins_with_elicitation_handler.yaml	Adds replay snapshot for multi-client commands/elicitation coverage.
test/snapshots/multi_client_commands_elicitation/capabilities_changed_fires_when_elicitation_provider_disconnects.yaml	Adds replay snapshot for multi-client commands/elicitation coverage.
test/snapshots/mcp_and_agents/should_handle_multiple_mcp_servers.yaml	Adds replay snapshot for MCP/agents E2E coverage.
test/snapshots/mcp_and_agents/should_handle_multiple_custom_agents.yaml	Adds replay snapshot for MCP/agents E2E coverage.
test/snapshots/mcp_and_agents/should_handle_custom_agent_with_tools_configuration.yaml	Adds replay snapshot for MCP/agents E2E coverage.
test/snapshots/mcp_and_agents/should_handle_custom_agent_with_mcp_servers.yaml	Adds replay snapshot for MCP/agents E2E coverage.
test/snapshots/elicitation/should_report_elicitation_capability_based_on_handler_presence.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/session_without_elicitationhandler_creates_successfully.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/sends_requestelicitation_when_handler_provided.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/select_returns_selected_option.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/input_returns_freeform_value.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/elicitation_throws_when_capability_is_missing.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/elicitation_returns_all_action_shapes.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/defaults_capabilities_when_not_provided.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/confirm_returns_true_when_handler_accepts.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/elicitation/confirm_returns_false_when_handler_declines.yaml	Adds replay snapshot for elicitation E2E coverage.
test/snapshots/commands/session_with_no_commands_creates_successfully.yaml	Adds replay snapshot for commands E2E coverage.
test/snapshots/commands/session_with_commands_resumes_successfully.yaml	Adds replay snapshot for commands E2E coverage.
test/snapshots/commands/session_with_commands_creates_successfully.yaml	Adds replay snapshot for commands E2E coverage.
test/snapshots/client/should_stop_client_with_active_session.yaml	Adds replay snapshot for client lifecycle E2E coverage.
test/snapshots/client/should_start_ping_and_stop_tcp_client.yaml	Adds replay snapshot for client connectivity E2E coverage.
test/snapshots/client/should_start_ping_and_stop_stdio_client.yaml	Adds replay snapshot for client connectivity E2E coverage.
test/snapshots/client/should_list_models_when_authenticated.yaml	Adds replay snapshot for list-models E2E coverage.
test/snapshots/client/should_get_status.yaml	Adds replay snapshot for status E2E coverage.
test/snapshots/client/should_get_authenticated_status.yaml	Adds replay snapshot for auth-status E2E coverage.
test/snapshots/client/should_force_stop_client.yaml	Adds replay snapshot for force-stop E2E coverage.
test/snapshots/client/listmodels_withcustomhandler_callshandler.yaml	Adds replay snapshot for list-models handler E2E coverage.
test/snapshots/client_options/should_listen_on_configured_tcp_port.yaml	Adds replay snapshot for client-options E2E coverage.
rust/tests/e2e/telemetry.rs	Adds replay-backed E2E validating file telemetry export contents/spans.
rust/tests/e2e/system_message_transform.rs	Adds replay-backed E2Es for system-message transform callbacks and overrides interplay.
rust/tests/e2e/suspend.rs	Adds suspend/resume E2Es and config-level assertions for continue-pending-work.
rust/tests/e2e/skills.rs	Adds skill discovery/disable/agent-skill E2Es plus a skipped resume-skill placeholder.
rust/tests/e2e/session_lifecycle.rs	Adds E2Es for listing sessions, deleting sessions, and event isolation across sessions.
rust/tests/e2e/rpc_tasks_and_handlers.rs	Adds tasks/handlers RPC E2Es plus placeholder coverage for background tasks.
rust/tests/e2e/rpc_shell_and_fleet.rs	Adds shell RPC E2Es plus a placeholder Fleet test.
rust/tests/e2e/rpc_server.rs	Adds typed server-scoped RPC E2Es (ping/models/tools/quota/mcp/skills).
rust/tests/e2e/rpc_mcp_config.rs	Adds MCP config RPC E2Es including HTTP OAuth config round-tripping.
rust/tests/e2e/rpc_extensions_loaded.rs	Adds extension-loaded tests (currently structured as local type assertions).
rust/tests/e2e/rpc_event_side_effects.rs	Adds E2Es validating events emitted as side effects of RPC calls (mode/plan/title/workspace/history).
rust/tests/e2e/rpc_agent.rs	Adds agent RPC E2Es for list/select/deselect/reload and event emissions.
rust/tests/e2e/per_session_auth.rs	Adds E2Es covering client-vs-session token behavior and invalid-token failures.
rust/tests/e2e/multi_turn.rs	Adds multi-turn E2Es validating tool sequencing and cross-turn file effects.
rust/tests/e2e/multi_client_commands_elicitation.rs	Adds multi-client E2Es validating commands-changed and elicitation-capability changes across clients.
rust/tests/e2e/mode_handlers.rs	Adds E2Es covering exit-plan-mode and auto-mode-switch handler flows (including rate-limit path).
rust/tests/e2e/hooks.rs	Adds hook E2Es for pre/post tool-use hooks and deny behavior.
rust/tests/e2e/error_resilience.rs	Adds E2Es for disconnected-session error behavior and abort idempotency.
rust/tests/e2e/compaction.rs	Adds E2Es for infinite-session compaction events and disabled behavior.
rust/tests/e2e/commands.rs	Adds E2Es for commands on create/resume plus unit-ish cloning/property checks.
rust/tests/e2e/client.rs	Adds client start/stop/ping/status/auth/list-models E2Es and list-models handler caching tests.
rust/tests/e2e/client_options.rs	Adds client-options E2Es plus several option-composition/serialization checks.
rust/tests/e2e/client_lifecycle.rs	Adds E2Es for lifecycle subscriptions and session created/updated/deleted lifecycle events.
rust/tests/e2e/client_api.rs	Adds E2Es for delete session, last session id, and foreground session behavior.
rust/tests/e2e/builtin_tools.rs	Adds E2Es validating built-in tool behaviors (shell/file/search/edit/create/find).
rust/tests/e2e/ask_user.rs	Adds E2Es for ask_user tool bridging via SessionHandler user-input callback.
rust/tests/e2e/abort.rs	Adds E2Es for abort during streaming and abort during tool execution, validating recovery.
rust/tests/e2e.rs	Registers the expanded Rust E2E module set under the test-support feature.
rust/src/types.rs	Extends ResumeSessionConfig with reasoning_effort and builder/debug support.
rust/src/session.rs	Refactors create/resume to pre-register sessions, enforce session-id consistency, and ensure cancellation-safe unregister cleanup.
rust/src/lib.rs	Adds cached list_models single-flight behavior, session-id generation helper, and resets model cache on shutdown paths.
rust/src/jsonrpc.rs	Tracks JSON-RPC read/write tasks and adds force_close() for force-stop semantics.

Copilot's findings

Comments suppressed due to low confidence (2)

rust/tests/e2e/suspend.rs:104

• should_reject_pending_external_tool_when_suspending is currently a config-field assertion and doesn't exercise suspend behavior or validate the CLI's handling of pending external tools. Consider adding a replay-backed E2E flow (create pending external tool → suspend → assert rejection) or rename/move this to a unit/config test so it doesn't read as E2E coverage.
  rust/tests/e2e/client_options.rs:260
• This test name says it should throw, but it only constructs ClientOptions and asserts fields—no error is ever triggered. Consider either asserting the actual error behavior via Client::start or renaming/moving this as a pure options-serialization/unit test.

async fn should_throw_when_useloggedinuser_used_with_cliurl() {
    let options = ClientOptions::new()
        .with_transport(Transport::External {
            host: "localhost".to_string(),
            port: 12345,
        })
        .with_use_logged_in_user(true);

    assert!(matches!(options.transport, Transport::External { .. }));
    assert_eq!(options.use_logged_in_user, Some(true));
}

• Files reviewed: 161/161 changed files
• Comments generated: 5