fabro/lib/crates/fabro-cli/tests/it/cmd/run.rs

#![expect(
    clippy::disallowed_methods,
    reason = "integration tests stage fixtures with sync std::fs; test infrastructure, not Tokio-hot path"
)]

use fabro_auth::{AuthCredential, AuthDetails};
use fabro_config::Storage;
use fabro_model::Provider;
use fabro_test::{fabro_json_snapshot, fabro_snapshot, test_context};
use fabro_vault::{SecretType, Vault};
use httpmock::MockServer;
use serde_json::Value;

use super::support::{output_stderr, wait_for_event_names};
use crate::support::{run_output_filters, unique_run_id};

fn run_status_response(run_id: &str, status: &str) -> serde_json::Value {
    let status = match status {
        "submitted" => serde_json::json!({ "kind": "submitted" }),
        "queued" => serde_json::json!({ "kind": "queued" }),
        other => panic!("unsupported test status {other:?}"),
    };
    serde_json::json!({
        "id": run_id,
        "status": status,
        "created_at": "2026-04-05T12:00:00Z"
    })
}

fn remote_run_state_response() -> serde_json::Value {
    serde_json::json!({
        "spec": null,
        "graph_source": null,
        "start": null,
        "status": null,
        "checkpoint": {
            "timestamp": "2026-04-05T12:00:01Z",
            "current_node": "exit",
            "completed_nodes": ["report"],
            "node_retries": {},
            "context_values": {
                "response.report": "Remote output"
            },
            "node_outcomes": {},
            "next_node_id": null,
            "git_commit_sha": null,
            "loop_failure_signatures": {},
            "restart_failure_signatures": {},
            "node_visits": {}
        },
        "checkpoints": [],
        "conclusion": {
            "timestamp": "2026-04-05T12:00:01Z",
            "status": "succeeded",
            "duration_ms": 12,
            "stages": [],
            "billing": null,
            "total_retries": 0,
        },
        "retro": null,
        "retro_prompt": null,
        "retro_response": null,
        "sandbox": null,
        "final_patch": null,
        "pull_request": null,
        "stages": {}
    })
}

fn run_completed_event(run_id: &str) -> serde_json::Value {
    serde_json::json!({
        "seq": 1,
        "event": "run.completed",
        "id": "evt-run-completed",
        "run_id": run_id,
        "ts": "2026-04-05T12:00:01Z",
        "properties": {
            "duration_ms": 12,
            "artifact_count": 0,
            "status": "succeeded",
            "reason": "completed"
        }
    })
}

fn seed_anthropic_vault(storage_dir: &std::path::Path, base_url: &str) {
    let mut vault =
        Vault::load(Storage::new(storage_dir).secrets_path()).expect("test vault should load");
    vault
        .set(
            "anthropic",
            &serde_json::to_string(&AuthCredential {
                provider: Provider::Anthropic,
                details:  AuthDetails::ApiKey {
                    key: "vault-anthropic-key".to_string(),
                },
            })
            .expect("Anthropic test credential should serialize"),
            SecretType::Credential,
            None,
        )
        .expect("Anthropic credential should store in test vault");
    vault
        .set(
            "ANTHROPIC_BASE_URL",
            base_url,
            SecretType::Environment,
            None,
        )
        .expect("Anthropic base URL should store in test vault");
}

fn run_running_event(run_id: &str, seq: u32) -> serde_json::Value {
    serde_json::json!({
        "seq": seq,
        "event": "run.running",
        "id": format!("evt-run-running-{seq}"),
        "run_id": run_id,
        "ts": "2026-04-05T12:00:00Z",
        "properties": {}
    })
}

#[test]
fn help() {
    let context = test_context!();
    let mut cmd = context.run_cmd();
    cmd.arg("--help");
    fabro_snapshot!(context.filters(), cmd, @"
    success: true
    exit_code: 0
    ----- stdout -----
    Launch a workflow run

    Usage: fabro run [OPTIONS] <WORKFLOW>

    Arguments:
      <WORKFLOW>  Path to a .fabro workflow file or .toml task config

    Options:
          --json                   Output as JSON [env: FABRO_JSON=]
          --server <SERVER>        Fabro server target: http(s) URL or absolute Unix socket path [env: FABRO_SERVER=]
          --debug                  Enable DEBUG-level logging (default is INFO) [env: FABRO_DEBUG=]
          --dry-run                Execute with simulated LLM backend
          --auto-approve           Auto-approve all human gates
          --no-upgrade-check       Disable automatic upgrade check [env: FABRO_NO_UPGRADE_CHECK=true]
          --goal <GOAL>            Override the workflow goal (available as {{ goal }} in prompts)
          --quiet                  Suppress non-essential output [env: FABRO_QUIET=]
          --goal-file <GOAL_FILE>  Read the workflow goal from a file
          --model <MODEL>          Override default LLM model
          --provider <PROVIDER>    Override default LLM provider
      -v, --verbose                Enable verbose output
          --sandbox <SANDBOX>      Sandbox for agent tools [possible values: local, docker, daytona]
          --in-place               Run directly in the source checkout without git checkpoints
          --label <KEY=VALUE>      Attach a label to this run (repeatable, format: KEY=VALUE)
          --no-retro               Skip retro generation after the run
          --preserve-sandbox       Keep the sandbox alive after the run finishes (for debugging)
      -d, --detach                 Run the workflow in the background and print the run ID
      -h, --help                   Print help
    ----- stderr -----
    ");
}

#[test]
fn detach_uses_explicit_server_target_and_prints_remote_run_id() {
    let context = test_context!();
    let server = MockServer::start();
    let run_id = unique_run_id();
    let create_mock = server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(201)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "submitted").to_string());
    });
    let start_mock = server.mock(|when, then| {
        when.method("POST")
            .path(format!("/api/v1/runs/{run_id}/start"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "queued").to_string());
    });

    let workflow = context.install_fixture("simple.fabro");
    let output = context
        .run_cmd()
        .args([
            "--server",
            &format!("{}/api/v1", server.base_url()),
            "--detach",
            "--dry-run",
            "--auto-approve",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    create_mock.assert();
    start_mock.assert();
    assert_eq!(output_stderr(&output), "");
    assert_eq!(
        String::from_utf8_lossy(&output.stdout).trim(),
        run_id.as_str()
    );
}

#[test]
fn detach_uses_configured_server_target_without_server_flag() {
    let context = test_context!();
    let server = MockServer::start();
    let run_id = unique_run_id();
    let create_mock = server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(201)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "submitted").to_string());
    });
    let start_mock = server.mock(|when, then| {
        when.method("POST")
            .path(format!("/api/v1/runs/{run_id}/start"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "queued").to_string());
    });
    context.set_http_target(&server.base_url());

    let workflow = context.install_fixture("simple.fabro");
    let output = context
        .run_cmd()
        .args([
            "--detach",
            "--dry-run",
            "--auto-approve",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    create_mock.assert();
    start_mock.assert();
    assert_eq!(
        String::from_utf8_lossy(&output.stdout).trim(),
        run_id.as_str()
    );
}

#[test]
fn run_uses_vault_credentials_for_worker_execution() {
    let mut context = test_context!();
    context.write_home(
        ".fabro/settings.toml",
        "[server.auth]\nmethods = [\"dev-token\"]\n",
    );
    context.isolated_server();
    let run_id = unique_run_id();
    let llm_server = MockServer::start();
    seed_anthropic_vault(
        &context.storage_dir,
        &format!("{}/v1", llm_server.base_url()),
    );

    let llm_mock = llm_server.mock(|when, then| {
        when.method("POST")
            .path("/v1/messages")
            .header("x-api-key", "vault-anthropic-key");
        then.status(200)
            .header("Content-Type", "application/json")
            .body(
                serde_json::json!({
                    "id": "msg_test_123",
                    "model": "claude-haiku-4-5",
                    "content": [
                        {
                            "type": "text",
                            "text": "Hello from vault"
                        }
                    ],
                    "stop_reason": "end_turn",
                    "usage": {
                        "input_tokens": 12,
                        "output_tokens": 4
                    }
                })
                .to_string(),
            );
    });

    context.write_temp(
        "vault_worker_llm.fabro",
        "\
digraph VaultWorkerLlm {
  graph [goal=\"Use a vault-backed Anthropic credential\"]
  rankdir=LR

  start [shape=Mdiamond, label=\"Start\"]
  exit  [shape=Msquare, label=\"Exit\"]
  draft [shape=tab, label=\"Draft\", prompt=\"Write a short greeting.\"]

  start -> draft -> exit
}
",
    );

    let output = context
        .run_cmd()
        .env_remove("ANTHROPIC_API_KEY")
        .env_remove("ANTHROPIC_BASE_URL")
        .env_remove("OPENAI_API_KEY")
        .env_remove("OPENAI_BASE_URL")
        .env_remove("GEMINI_API_KEY")
        .args([
            "--run-id",
            run_id.as_str(),
            "--auto-approve",
            "--no-retro",
            "--sandbox",
            "local",
            "--provider",
            "anthropic",
            "--model",
            "claude-haiku-4-5",
            context
                .temp_dir
                .join("vault_worker_llm.fabro")
                .to_str()
                .unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );

    llm_mock.assert();
    wait_for_event_names(&context.find_run_dir(&run_id), &["run.completed"]);
}

#[test]
fn detach_rejects_storage_dir_flag() {
    let context = test_context!();
    let workflow = context.install_fixture("simple.fabro");
    let output = context
        .run_cmd()
        .args([
            "--storage-dir",
            "/tmp/fabro-run",
            "--detach",
            "--dry-run",
            "--auto-approve",
            "--no-retro",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        !output.status.success(),
        "command should reject --storage-dir"
    );
    let stderr = String::from_utf8_lossy(&output.stderr);
    assert!(stderr.contains("unexpected argument '--storage-dir'"));
}

#[test]
fn detach_cli_server_target_overrides_configured_server_target() {
    let context = test_context!();
    let config_server = MockServer::start();
    let config_create = config_server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(500)
            .body("configured-server-should-not-be-used");
    });
    let config_start = config_server.mock(|when, then| {
        when.method("POST").path_includes("/api/v1/runs/");
        then.status(500)
            .body("configured-server-should-not-be-used");
    });
    let cli_server = MockServer::start();
    let run_id = unique_run_id();
    let cli_create = cli_server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(201)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "submitted").to_string());
    });
    let cli_start = cli_server.mock(|when, then| {
        when.method("POST")
            .path(format!("/api/v1/runs/{run_id}/start"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "queued").to_string());
    });
    context.set_http_target(&config_server.base_url());

    let workflow = context.install_fixture("simple.fabro");
    let output = context
        .run_cmd()
        .args([
            "--server",
            &format!("{}/api/v1", cli_server.base_url()),
            "--detach",
            "--dry-run",
            "--auto-approve",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    cli_create.assert();
    cli_start.assert();
    config_create.assert_calls(0);
    config_start.assert_calls(0);
    assert_eq!(
        String::from_utf8_lossy(&output.stdout).trim(),
        run_id.as_str()
    );
}

#[test]
fn remote_foreground_run_consumes_paginated_events_and_prints_server_backed_summary() {
    let context = test_context!();
    let server = MockServer::start();
    let run_id = unique_run_id();
    let preflight = server.mock(|when, then| {
        when.method("POST").path("/api/v1/preflight");
        then.status(500)
            .header("Content-Type", "application/json")
            .body(serde_json::json!({ "error": "preflight should not run" }).to_string());
    });
    server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(201)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "submitted").to_string());
    });
    server.mock(|when, then| {
        when.method("POST")
            .path(format!("/api/v1/runs/{run_id}/start"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(run_status_response(run_id.as_str(), "queued").to_string());
    });
    let first_page = server.mock(|when, then| {
        when.method("GET")
            .path(format!("/api/v1/runs/{run_id}/events"))
            .query_param_missing("since_seq");
        then.status(200)
            .header("Content-Type", "application/json")
            .body(
                serde_json::json!({
                    "data": [run_running_event(run_id.as_str(), 1)],
                    "meta": { "has_more": true }
                })
                .to_string(),
            );
    });
    let second_page = server.mock(|when, then| {
        when.method("GET")
            .path(format!("/api/v1/runs/{run_id}/events"))
            .query_param("since_seq", "2");
        then.status(200)
            .header("Content-Type", "application/json")
            .body(
                serde_json::json!({
                    "data": [run_completed_event(run_id.as_str())],
                    "meta": { "has_more": false }
                })
                .to_string(),
            );
    });
    server.mock(|when, then| {
        when.method("GET")
            .path(format!("/api/v1/runs/{run_id}/questions"))
            .query_param("page[limit]", "100")
            .query_param("page[offset]", "0");
        then.status(200)
            .header("Content-Type", "application/json")
            .body(
                serde_json::json!({
                    "data": [],
                    "meta": { "has_more": false }
                })
                .to_string(),
            );
    });
    server.mock(|when, then| {
        when.method("GET")
            .path(format!("/api/v1/runs/{run_id}/state"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(remote_run_state_response().to_string());
    });
    server.mock(|when, then| {
        when.method("GET")
            .path(format!("/api/v1/runs/{run_id}/artifacts"));
        then.status(200)
            .header("Content-Type", "application/json")
            .body(serde_json::json!({ "data": [] }).to_string());
    });

    let workflow = context.install_fixture("simple.fabro");
    let output = context
        .run_cmd()
        .args([
            "--server",
            &format!("{}/api/v1", server.base_url()),
            "--dry-run",
            "--auto-approve",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    preflight.assert_calls(0);
    first_page.assert();
    second_page.assert();

    let stderr = output_stderr(&output);
    assert!(stderr.contains("=== Run Result ==="), "{stderr}");
    assert!(stderr.contains("Remote output"), "{stderr}");
    assert_eq!(
        stderr
            .lines()
            .filter(|line| line.trim_start().starts_with("Run:"))
            .count(),
        2,
        "{stderr}"
    );
    assert!(!stderr.contains("=== Artifacts ==="), "{stderr}");
}

#[test]
fn foreground_run_rejects_invalid_workflow_before_creating_remote_run() {
    let context = test_context!();
    let server = MockServer::start();
    let create = server.mock(|when, then| {
        when.method("POST").path("/api/v1/runs");
        then.status(500)
            .header("Content-Type", "application/json")
            .body(serde_json::json!({ "error": "run should not be created" }).to_string());
    });

    let workflow = context.install_fixture("invalid.fabro");
    let output = context
        .run_cmd()
        .args([
            "--server",
            &format!("{}/api/v1", server.base_url()),
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        !output.status.success(),
        "invalid run should fail:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    create.assert_calls(0);

    let stderr = output_stderr(&output);
    assert!(stderr.contains("Workflow: Invalid"), "{stderr}");
    assert!(
        stderr.contains("Pipeline must have exactly one start node"),
        "{stderr}"
    );
    assert!(stderr.contains("Validation failed"), "{stderr}");
}

#[test]
fn local_foreground_run_prints_artifact_paths_from_server_artifact_list() {
    let context = test_context!();
    let run_id = unique_run_id();
    let workspace_dir = context.temp_dir.join("artifact-summary");
    context.write_temp(
        "artifact-summary/workflow.fabro",
        r#"digraph ArtifactSummary {
  graph [goal="Show stored artifacts"]
  start [shape=Mdiamond]
  exit [shape=Msquare]
  create_assets [shape=parallelogram, script="mkdir -p assets/shared && printf one > assets/shared/report.txt"]
  start -> create_assets -> exit
}
"#,
    );
    context.write_temp(
        "artifact-summary/run.toml",
        r#"_version = 1

[workflow]
graph = "workflow.fabro"

[run]
goal = "Show stored artifacts"

[run.sandbox]
provider = "local"
preserve = true

[run.sandbox.local]
worktree_mode = "never"

[run.artifacts]
include = ["assets/**"]
"#,
    );

    let output = context
        .run_cmd()
        .current_dir(&workspace_dir)
        .env("OPENAI_API_KEY", "test")
        .args([
            "--run-id",
            run_id.as_str(),
            "--auto-approve",
            "--no-retro",
            "--sandbox",
            "local",
            "--provider",
            "openai",
            "run.toml",
        ])
        .output()
        .expect("command should execute");

    assert!(
        output.status.success(),
        "command failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );

    let stderr = output_stderr(&output);
    assert!(stderr.contains("=== Artifacts ==="), "{stderr}");
    assert!(!stderr.contains("cache/artifacts"), "{stderr}");
    assert!(stderr.contains("create_assets"), "{stderr}");
    assert!(stderr.contains("assets/shared/report.txt"), "{stderr}");
    assert!(stderr.contains("fabro artifact cp"), "{stderr}");
}

#[test]
fn dry_run_simple() {
    let context = test_context!();
    let workflow = context.install_fixture("simple.fabro");
    let mut cmd = context.run_cmd();
    cmd.args(["--dry-run", "--auto-approve"]);
    cmd.arg(&workflow);
    fabro_snapshot!(run_output_filters(&context), cmd, @"
    success: true
    exit_code: 0
    ----- stdout -----
    ----- stderr -----
    Workflow: Simple (4 nodes, 3 edges)
    Graph: [GRAPH_PATH]
    Goal: Run tests and report results

        Run: [ULID]
        Web UI: http://localhost:3000/runs/[ULID]
        Warning: Worktree mode `always` requested but no Git repository was found; running without a worktree. [worktree_skipped_no_git]
        Sandbox: local (ready in [TIME])
        ✓ Start  [TIME]
        ✓ Run Tests  [TIME]
        ✓ Report  [TIME]
        ✓ Exit  [TIME]

    === Run Result ===
    Run:       [ULID]
    Status:    SUCCEEDED
    Duration:  [DURATION]

    === Output ===
    [Simulated] Response for stage: report
    ");
}

#[test]
fn dry_run_with_goal_file_reads_contents_into_goal() {
    // Regression test for the `--goal-file` flag that was previously
    // being silently ignored in the v2 path. The file content must end
    // up in the effective goal displayed in the workflow summary.
    let context = test_context!();

    let goal_dir = tempfile::tempdir().unwrap();
    let goal_path = goal_dir.path().join("goal.md");
    std::fs::write(&goal_path, "Ship the rate-limiting feature end to end.\n").unwrap();

    let workflow = context.install_fixture("simple.fabro");
    let mut cmd = context.run_cmd();
    cmd.args(["--dry-run", "--auto-approve", "--goal-file"]);
    cmd.arg(&goal_path);
    cmd.arg(&workflow);

    let output = cmd.output().expect("run command should execute");
    assert!(
        output.status.success(),
        "run should succeed:\nstderr:\n{}",
        String::from_utf8_lossy(&output.stderr)
    );
    let stderr = String::from_utf8_lossy(&output.stderr);
    assert!(
        stderr.contains("Ship the rate-limiting feature end to end."),
        "goal file content should appear in workflow summary, got:\n{stderr}"
    );
}

#[test]
fn dry_run_rejects_goal_and_goal_file_together() {
    // clap `conflicts_with` must fire when both flags are supplied.
    let context = test_context!();

    let goal_dir = tempfile::tempdir().unwrap();
    let goal_path = goal_dir.path().join("goal.md");
    std::fs::write(&goal_path, "never read").unwrap();

    let workflow = context.install_fixture("simple.fabro");
    let mut cmd = context.run_cmd();
    cmd.args(["--dry-run", "--goal", "inline override", "--goal-file"]);
    cmd.arg(&goal_path);
    cmd.arg(&workflow);
    let output = cmd.output().expect("run command should execute");
    assert!(
        !output.status.success(),
        "run should fail when --goal and --goal-file are both set"
    );
    let stderr = String::from_utf8_lossy(&output.stderr);
    assert!(
        stderr.contains("cannot be used with")
            || stderr.contains("conflict")
            || stderr.to_lowercase().contains("mutually exclusive"),
        "expected conflicts_with error, got:\n{stderr}"
    );
}

#[test]
fn dry_run_persists_event_history_in_store() {
    let context = test_context!();
    context.ensure_home_server_auth_methods();
    let run_id = unique_run_id();
    let workflow = context.install_fixture("simple.fabro");

    context
        .command()
        .args([
            "run",
            "--dry-run",
            "--auto-approve",
            "--sandbox",
            "local",
            "--run-id",
            run_id.as_str(),
            workflow.to_str().unwrap(),
        ])
        .assert()
        .success();

    let run_dir = context.find_run_dir(&run_id);
    wait_for_event_names(&run_dir, &["run.completed", "sandbox.cleanup.completed"]);
    let output = context
        .command()
        .args(["logs", &run_id])
        .output()
        .expect("logs command should execute");
    assert!(
        output.status.success(),
        "logs failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );
    let progress: Vec<Value> = String::from_utf8(output.stdout)
        .expect("stdout should be UTF-8")
        .lines()
        .filter(|line| !line.trim().is_empty())
        .map(|line| serde_json::from_str(line).expect("logs output should be JSONL"))
        .collect();
    assert!(
        !progress.is_empty(),
        "store-backed event history should have at least one line"
    );
    assert_eq!(
        progress.first().and_then(|event| event["event"].as_str()),
        Some("run.created")
    );
    assert_eq!(
        progress
            .first()
            .and_then(|event| event.pointer("/properties/settings/run/execution/approval"))
            .and_then(Value::as_str),
        Some("auto")
    );
    assert!(
        progress
            .iter()
            .any(|event| event["event"].as_str() == Some("run.completed")),
        "store-backed event history should include run.completed"
    );
    assert_eq!(
        progress.last().and_then(|event| event["event"].as_str()),
        Some("sandbox.cleanup.completed")
    );

    let tail_output = context
        .command()
        .args(["logs", "--tail", "1", &run_id])
        .output()
        .expect("tail logs command should execute");
    assert!(
        tail_output.status.success(),
        "tail logs failed:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&tail_output.stdout),
        String::from_utf8_lossy(&tail_output.stderr)
    );
    let live_content: Value = String::from_utf8(tail_output.stdout)
        .expect("stdout should be UTF-8")
        .lines()
        .find(|line| !line.trim().is_empty())
        .map(|line| serde_json::from_str(line).expect("tail logs output should be JSON"))
        .expect("tail logs should include the latest event");
    fabro_json_snapshot!(context, &live_content, @r#"
    {
      "actor": {
        "kind": "worker",
        "run_id": "[ULID]"
      },
      "event": "sandbox.cleanup.completed",
      "id": "[EVENT_ID]",
      "properties": {
        "duration_ms": "[DURATION_MS]",
        "provider": "local"
      },
      "run_id": "[ULID]",
      "ts": "[TIMESTAMP]"
    }
    "#);

    assert_eq!(live_content, *progress.last().unwrap());
}

#[test]
fn run_id_passthrough_uses_provided_ulid() {
    let context = test_context!();
    context.ensure_home_server_auth_methods();
    let run_id = unique_run_id();
    let workflow = context.install_fixture("simple.fabro");

    context
        .command()
        .args([
            "run",
            "--dry-run",
            "--auto-approve",
            "--run-id",
            run_id.as_str(),
            workflow.to_str().unwrap(),
        ])
        .assert()
        .success();

    context.find_run_dir(&run_id);
}

#[test]
fn json_run_requires_manual_input_for_human_gates_without_auto_approve() {
    let context = test_context!();
    context.ensure_home_server_auth_methods();
    let workflow = context.temp_dir.join("human-gate.fabro");
    context.write_temp(
        "human-gate.fabro",
        r#"digraph HumanGate {
  graph [goal="Require explicit approval before continuing"]
  start [shape=Mdiamond, label="Start"]
  exit  [shape=Msquare, label="Exit"]
  approve [shape=hexagon, label="Approve?"]
  ship   [shape=parallelogram, script="echo shipped"]
  revise [shape=parallelogram, script="echo revised"]
  start -> approve
  approve -> ship   [label="[A] Approve"]
  approve -> revise [label="[R] Revise"]
  ship -> exit
  revise -> exit
}
"#,
    );

    let output = context
        .command()
        .args([
            "--json",
            "run",
            "--sandbox",
            "local",
            "--no-retro",
            workflow.to_str().unwrap(),
        ])
        .output()
        .expect("command should execute");

    assert!(
        !output.status.success(),
        "command unexpectedly succeeded:\nstdout:\n{}\nstderr:\n{}",
        String::from_utf8_lossy(&output.stdout),
        String::from_utf8_lossy(&output.stderr)
    );

    let stderr = String::from_utf8(output.stderr).expect("stderr should be UTF-8");
    assert!(
        stderr.contains("This run is waiting for human input, but --json is non-interactive."),
        "stderr should explain the non-interactive interview failure:\n{stderr}"
    );

    let progress: Vec<Value> = String::from_utf8(output.stdout)
        .expect("stdout should be UTF-8")
        .lines()
        .filter(|line| !line.trim().is_empty())
        .map(|line| serde_json::from_str(line).expect("run JSON output should be JSONL"))
        .collect();

    assert!(
        progress
            .iter()
            .any(|event| event.get("event") == Some(&Value::String("interview.started".into()))),
        "stdout should include the interview start event:\n{}",
        serde_json::to_string_pretty(&progress).unwrap()
    );
}

#[test]
fn detach_prints_ulid_and_exits() {
    let context = test_context!();
    let workflow = context.install_fixture("simple.fabro");
    let mut cmd = context.run_cmd();
    cmd.args([
        "--detach",
        "--dry-run",
        "--auto-approve",
        workflow.to_str().unwrap(),
    ]);
    fabro_snapshot!(context.filters(), cmd, @"
    success: true
    exit_code: 0
    ----- stdout -----
    [ULID]
    ----- stderr -----
    ");
}

#[test]
fn detach_creates_run_dir_with_detach_log() {
    let context = test_context!();
    let run_id = unique_run_id();
    let workflow = context.install_fixture("simple.fabro");

    context
        .run_cmd()
        .args([
            "--detach",
            "--dry-run",
            "--auto-approve",
            "--run-id",
            run_id.as_str(),
            workflow.to_str().unwrap(),
        ])
        .assert()
        .success();

    let run_dir = context.find_run_dir(&run_id);
    fabro_json_snapshot!(
        context,
        serde_json::json!({
            "run_dir": run_dir,
            "launcher_log_exists": context.storage_dir.join("launchers").join(format!("{run_id}.log")).exists(),
        }),
        @r#"
    {
      "run_dir": "[RUN_DIR]",
      "launcher_log_exists": false
    }
    "#
    );
}