hyperhive/hive-c0re/src/server.rs

use std::path::Path;
use std::sync::Arc;

use anyhow::{Context, Result};
use hive_host_sock::{HostRequest, HostResponse, LifecycleScope};
use hive_priv_sock::{InfraAction, InfraContainer};
use tokio::io::{AsyncBufReadExt, AsyncWriteExt, BufReader};
use tokio::net::{UnixListener, UnixStream};

use crate::actions;
use crate::coordinator::Coordinator;
use crate::lifecycle;

pub async fn serve(socket: &Path, coord: Arc<Coordinator>) -> Result<()> {
    // Prefer a socket passed by systemd socket-activation (LISTEN_FDS).
    // When running under a `.socket` unit, systemd has already created,
    // bound, and chmod-ed the socket for us — we just accept on it.
    // Fall back to the traditional bind path when not socket-activated
    // (direct invocation, dev, tests).
    let listener = {
        let mut listenfd = listenfd::ListenFd::from_env();
        if let Some(std_listener) = listenfd
            .take_unix_listener(0)
            .context("take socket-activated unix listener")?
        {
            std_listener.set_nonblocking(true)?;
            UnixListener::from_std(std_listener)
                .context("convert socket-activated listener to tokio")?
        } else {
            // Standalone: create parent dir, remove any stale socket, bind.
            if let Some(parent) = socket.parent() {
                std::fs::create_dir_all(parent)
                    .with_context(|| format!("create socket parent {}", parent.display()))?;
            }
            if socket.exists() {
                std::fs::remove_file(socket).context("remove stale socket")?;
            }
            UnixListener::bind(socket)
                .with_context(|| format!("bind admin socket {}", socket.display()))?
        }
    };
    tracing::info!(socket = %socket.display(), hyperhive_flake = %coord.hyperhive_flake, "hive-c0re admin listening");

    loop {
        let (stream, _) = listener.accept().await.context("accept connection")?;
        let coord = coord.clone();
        tokio::spawn(async move {
            if let Err(e) = handle(stream, coord).await {
                tracing::warn!(error = ?e, "connection failed");
            }
        });
    }
}

async fn handle(stream: UnixStream, coord: Arc<Coordinator>) -> Result<()> {
    let (read, mut write) = stream.into_split();
    let mut reader = BufReader::new(read);
    let mut line = String::new();

    loop {
        line.clear();
        let n = reader.read_line(&mut line).await?;
        if n == 0 {
            return Ok(());
        }
        let resp = match serde_json::from_str::<HostRequest>(line.trim()) {
            Ok(req) => dispatch(&req, coord.clone()).await,
            Err(e) => HostResponse::error(format!("parse error: {e}")),
        };
        let mut payload = serde_json::to_string(&resp)?;
        payload.push('\n');
        write.write_all(payload.as_bytes()).await?;
        write.flush().await?;
    }
}

async fn dispatch(req: &HostRequest, coord: Arc<Coordinator>) -> HostResponse {
    let result: anyhow::Result<HostResponse> = async {
        Ok(match req {
            HostRequest::Spawn { name } => handle_spawn(&coord, name).await?,
            HostRequest::RequestSpawn { name } => {
                tracing::info!(%name, "request_spawn");
                let id = coord.approvals.submit_kind(
                    name,
                    hive_sh4re::ApprovalKind::Spawn,
                    "",
                    None,
                    "operator",
                    None,
                )?;
                tracing::info!(%id, %name, "spawn approval queued");
                HostResponse::success()
            }
            HostRequest::Kill { name } => submit_single(&coord, name, Verb::Kill),
            HostRequest::Restart { name } => submit_single(&coord, name, Verb::Restart),
            HostRequest::RestartAll => handle_restart_all(&coord).await?,
            HostRequest::RestartScoped { scope, graceful } => {
                handle_restart_scoped(&coord, scope, *graceful).await?
            }
            HostRequest::Stop { scope, graceful } => {
                // Resolve the scope to explicit container names at the entry
                // point, then operate on names — never pass the bare "all
                // agents" flag deeper (it'd force every consumer, incl. the
                // graceful-stop queue, to re-expand it).
                let agents = scoped_agents(scope).await?;
                let infra = scoped_infra(scope);
                // On a broad stop, remember which agents were actually
                // running so a later broad `start` restores only those
                // (not every configured container). A targeted `--agent`
                // stop must not redefine the restore set.
                if is_broad_scope(scope) {
                    let mut running = Vec::new();
                    for a in &agents {
                        if lifecycle::is_running(a).await {
                            running.push(a.clone());
                        }
                    }
                    coord.set_last_stopped_running(running);
                }
                handle_stop(&coord, &agents, &infra, *graceful).await?
            }
            HostRequest::Start { scope } => {
                let mut agents = scoped_agents(scope).await?;
                // A broad start restores only the set recorded at the
                // last broad stop, if any. No record (cold "bring the
                // hive up", or a daemon restart since the stop) → start
                // all. Targeted `--agent` start is never filtered.
                if is_broad_scope(scope)
                    && let Some(prev) = coord.take_last_stopped_running()
                {
                    agents.retain(|a| prev.contains(a));
                }
                let infra = scoped_infra(scope);
                handle_start(&coord, &agents, &infra).await?
            }
            HostRequest::Destroy { name, purge } => {
                actions::destroy(&coord, name, *purge).await?;
                HostResponse::success()
            }
            HostRequest::Rebuild { name } => submit_single(&coord, name, Verb::Rebuild),
            HostRequest::QueueDag { id } => {
                // The polled DAG first, then its live fan-out children.
                let dags = coord
                    .job_queue
                    .snapshot()
                    .into_iter()
                    .filter(|d| d.id == *id || d.parent_id == Some(*id))
                    .collect();
                HostResponse::dags(dags)
            }
            HostRequest::List => HostResponse::list(lifecycle::list().await?),
            HostRequest::AgentStatus => handle_agent_status(&coord).await,
            // The hive domain + per-surface public URLs are injected into
            // c0re's service env by hive-c0re.nix; surface them so the
            // operator CLI can fill in this hive's own identity (the
            // federation peer-config block) and open the web surfaces
            // (`hivectl open`).
            HostRequest::Urls => HostResponse::urls(hive_urls()),
            HostRequest::Pending => HostResponse::pending(coord.approvals.pending()?),
            HostRequest::Approve { id } => {
                actions::approve(coord.clone(), *id).await?;
                HostResponse::success()
            }
            HostRequest::Deny { id } => {
                actions::deny(&coord, *id, None).await?;
                HostResponse::success()
            }
            HostRequest::SetParent { child, new_parent } => {
                tracing::info!(%child, ?new_parent, "set_parent");
                // `reparent_with_notify` wraps `topology::set_parent`
                // with the three notification messages + the
                // ContainerView rescan. Idempotent same-parent calls
                // skip both the messages and the disk write per the
                // topology fast-path.
                coord
                    .reparent_with_notify(child, new_parent.as_deref())
                    .await
                    .map_err(anyhow::Error::msg)?;
                HostResponse::success()
            }
            HostRequest::MatrixCreateUser { name, password } => {
                handle_matrix_create_user(name, password.as_deref()).await?
            }
            HostRequest::MatrixSyncAdmin => handle_matrix_sync_admin().await?,
            HostRequest::MatrixPromoteUser { name } => handle_matrix_promote_user(name).await?,
            HostRequest::MatrixResetPassword { name } => handle_matrix_reset_password(name).await?,
            HostRequest::MatrixInvite { user, room } => {
                handle_matrix_invite(user, room.as_deref()).await?
            }
        })
    }
    .await;
    match result {
        Ok(r) => r,
        Err(e) => HostResponse::error(format!("{e:#}")),
    }
}

/// Create + start the container for `name`, rolling back socket
/// registration and notifying the manager on failure.
async fn handle_spawn(coord: &Arc<Coordinator>, name: &str) -> Result<HostResponse> {
    tracing::info!(%name, "spawn");
    let agent_dir = crate::paths::agent_runtime_dir(name);
    let hive = coord.hive_env();
    let paths = Coordinator::agent_paths(name, agent_dir);
    // lifecycle::spawn creates the runtime dir internally before start.
    // MCP listener registration is event-driven: bind immediately on
    // success so the harness can connect on its first turn without
    // waiting for any poll interval.
    match lifecycle::spawn(name, &hive, &paths).await {
        Ok(()) => {
            if let Err(e) = coord.power.set(name, crate::power::Wanted::Up) {
                tracing::warn!(%name, error = ?e, "agent_power: set wanted=up failed");
            }
            // Bind the MCP listener now that the container is starting up.
            // The harness connects to this socket on its first turn.
            coord.register_agent(name)?;
            coord.notify_manager(&hive_sh4re::HelperEvent::Spawned {
                agent: name.to_owned(),
                ok: true,
                note: None,
                sha: None,
            });
            // Update tmpfiles.d so the new agent's dirs survive a reboot.
            tokio::spawn(lifecycle::sync_tmpfiles());
        }
        Err(e) => {
            // Spawn failed: register_agent was never called, so there is
            // nothing to unregister. Notify the manager and propagate.
            coord.notify_manager(&hive_sh4re::HelperEvent::Spawned {
                agent: name.to_owned(),
                ok: false,
                note: Some(format!("{e:#}")),
                sha: None,
            });
            return Err(e);
        }
    }
    Ok(HostResponse::success())
}

/// Collect per-agent status rows for `hivectl status` and the dashboard.
async fn handle_agent_status(coord: &Arc<Coordinator>) -> HostResponse {
    let rows = crate::container_view::build_all(coord)
        .await
        .into_iter()
        .map(|v| hive_sh4re::AgentStatusRow {
            name: v.name,
            running: v.running,
            needs_update: v.needs_update,
            needs_login: v.needs_login,
            deployed_sha: v.deployed_sha,
            pending_reminders: v.pending_reminders,
            parent: v.parent,
        })
        .collect();
    HostResponse::agent_statuses(rows)
}

// ---------------------------------------------------------------------------
// Matrix provisioning handlers
//
// The `hivectl matrix` subcommands used to run these in-process, which forced
// the standalone CLI to link the whole daemon crate (matrix-sdk, reqwest, …).
// They now run daemon-side over the host socket: the daemon already holds the
// register + admin tokens and the matrix creds dir. Each op returns the
// operator-facing lines hivectl used to `println!` in `HostResponse::messages`
// for the client to print verbatim.
// ---------------------------------------------------------------------------

/// Shared reqwest client for the matrix admin HTTP calls (30s timeout,
/// mirroring the old in-CLI client).
fn matrix_http_client() -> Result<reqwest::Client> {
    reqwest::Client::builder()
        .timeout(std::time::Duration::from_secs(30))
        .build()
        .context("build reqwest client")
}

/// True when `name` has a state dir under the agents root, i.e. it's a
/// managed agent rather than a bare (operator/human) matrix account.
fn agent_exists(name: &str) -> Result<bool> {
    crate::paths::agent_state_dir(name)
        .try_exists()
        .with_context(|| format!("check agent state dir for {name}"))
}

/// Guard: matrix provisioning needs the homeserver container running.
async fn require_matrix_present() -> Result<()> {
    if crate::matrix::is_present().await {
        return Ok(());
    }
    anyhow::bail!(
        "hive-matrix container not running — start it (services.hyperhive.matrix.enable = true) before provisioning matrix users"
    )
}

async fn handle_matrix_create_user(name: &str, password: Option<&str>) -> Result<HostResponse> {
    require_matrix_present().await?;
    let register_token =
        crate::matrix::ensure_register_token().context("read matrix register token")?;
    let client = matrix_http_client()?;
    let mut out = Vec::new();
    if agent_exists(name)? {
        if password.is_some() {
            // Agents auth by access_token, never by password — the
            // boot-sweep provisioning path doesn't accept one. Refuse
            // rather than silently dropping it.
            anyhow::bail!(
                "matrix create-user: a password is for non-agent (operator) accounts only; '{name}' is an agent which authenticates via access_token"
            );
        }
        crate::matrix::ensure_user_for(&client, name, &register_token)
            .await
            .with_context(|| format!("matrix create-user {name}"))?;
        let path = Coordinator::agent_notes_dir(name).join("matrix-token");
        out.push(format!("matrix: provisioned agent user '{name}'"));
        out.push(format!("token persisted at: {}", path.display()));
    } else {
        let effective_password = match password {
            Some(p) => p.to_owned(),
            None => crate::matrix::random_password().context("generate random matrix password")?,
        };
        let token = crate::matrix::provision_user_token(
            &client,
            name,
            &register_token,
            &effective_password,
        )
        .await
        .with_context(|| format!("matrix create-user {name}"))?;
        out.push(format!(
            "matrix: provisioned user '{name}' (not an agent — token not persisted)"
        ));
        out.push(format!("token: {token}"));
        if password.is_some() {
            out.push(
                "password: set as supplied — use it to log into a matrix web client".to_owned(),
            );
        } else {
            out.push(
                "password: random throwaway (not surfaced — pass --password or --password-stdin to set one you can use)".to_owned(),
            );
        }
    }
    Ok(HostResponse::messages(out))
}

async fn handle_matrix_sync_admin() -> Result<HostResponse> {
    require_matrix_present().await?;
    let register_token =
        crate::matrix::ensure_register_token().context("read matrix register token")?;
    let client = matrix_http_client()?;
    crate::matrix::ensure_admin_user(&client, &register_token)
        .await
        .context("matrix sync-admin")?;
    let path = crate::matrix::admin_token_path();
    Ok(HostResponse::messages(vec![
        format!(
            "matrix: hive admin user '@{}' provisioned",
            crate::matrix::HIVE_ADMIN_LOCALPART
        ),
        format!("token persisted at: {}", path.display()),
    ]))
}

async fn handle_matrix_promote_user(name: &str) -> Result<HostResponse> {
    require_matrix_present().await?;
    let admin_token = crate::matrix::read_admin_token()?;
    let client = matrix_http_client()?;
    let server_name = crate::matrix::discover_server_name(&client)
        .await
        .context("discover matrix server_name")?;
    crate::matrix::promote_user_to_admin(&client, &admin_token, name, &server_name)
        .await
        .with_context(|| format!("matrix promote-user {name}"))?;
    Ok(HostResponse::messages(vec![format!(
        "matrix: promoted @{name}:{server_name} to admin"
    )]))
}

async fn handle_matrix_invite(user: &str, room: Option<&str>) -> Result<HostResponse> {
    require_matrix_present().await?;
    let admin_token = crate::matrix::read_admin_token()?;
    let client = matrix_http_client()?;
    let server_name = crate::matrix::discover_server_name(&client)
        .await
        .context("discover matrix server_name")?;
    let room_id = crate::matrix::invite_user(&client, &admin_token, user, room, &server_name)
        .await
        .with_context(|| format!("matrix invite {user}"))?;
    let target = if user.starts_with('@') {
        user.to_owned()
    } else {
        format!("@{user}:{server_name}")
    };
    Ok(HostResponse::messages(vec![format!(
        "matrix: invited {target} to {room_id}"
    )]))
}

async fn handle_matrix_reset_password(name: &str) -> Result<HostResponse> {
    require_matrix_present().await?;
    let admin_token = crate::matrix::read_admin_token()?;
    let client = matrix_http_client()?;
    let server_name = crate::matrix::discover_server_name(&client)
        .await
        .context("discover matrix server_name")?;
    crate::matrix::reset_user_password(&client, &admin_token, name, &server_name)
        .await
        .with_context(|| format!("matrix reset-password {name}"))?;
    // Password is persisted by reset_user_password.
    let pw_path = crate::paths::matrix_creds_dir().join(format!("{name}-password"));
    Ok(HostResponse::messages(vec![
        format!("matrix: password for @{name}:{server_name} reset"),
        format!("password persisted at: {}", pw_path.display()),
        format!("next: hivectl matrix create-user {name}   # mints a fresh access token"),
    ]))
}

/// Single-agent queue verbs the admin socket exposes. Each submits the
/// matching DAG (persisting the `wanted` intent, serializing on the
/// agent's lease, with the transient/crash-watch suppression the old
/// direct lifecycle calls lacked) and returns the DAG id for the
/// client's wait loop.
#[derive(Clone, Copy)]
enum Verb {
    /// Stop DAG (`wanted = Offline`; Reconcile kills + unregisters +
    /// fires `Killed`).
    Kill,
    /// Restart DAG (`wanted = Up`; mechanical stop + reconcile-start).
    Restart,
    /// Rebuild DAG — the Swap tail owns the manager `Rebuilt` events +
    /// kick, so the CLI path can't drift from the dashboard's.
    Rebuild,
}

fn submit_single(coord: &Arc<Coordinator>, name: &str, verb: Verb) -> HostResponse {
    use crate::job_queue::{Source, submit};
    let id = match verb {
        Verb::Kill => {
            tracing::info!(%name, "kill");
            submit::stop(
                coord,
                name,
                Source::Manual,
                "manual kill via hivectl".to_owned(),
            )
        }
        Verb::Restart => {
            tracing::info!(%name, "restart");
            submit::restart(
                coord,
                name,
                Source::Manual,
                "manual restart via hivectl".to_owned(),
            )
        }
        Verb::Rebuild => {
            tracing::info!(%name, "rebuild");
            submit::rebuild(
                coord,
                name,
                Source::Manual,
                "manual rebuild via hivectl".to_owned(),
            )
        }
    };
    HostResponse::queued(vec![id])
}

/// Restart every container by submitting one restart DAG per agent —
/// each serializes on its own lease, so unrelated agents' restarts
/// overlap while nothing races an in-flight rebuild. Returns once all
/// are queued; per-agent results surface on the queue.
async fn handle_restart_all(coord: &Arc<Coordinator>) -> Result<HostResponse> {
    tracing::info!("restart-all");
    let agents = lifecycle::list().await?;
    let mut ok_agents: Vec<String> = Vec::new();
    let mut queued: Vec<u64> = Vec::new();
    for agent in &agents {
        let Some(logical) = agent.strip_prefix(lifecycle::AGENT_PREFIX) else {
            continue;
        };
        queued.push(crate::job_queue::submit::restart(
            coord,
            logical,
            crate::job_queue::Source::Manual,
            "manual restart via hivectl restart-all".to_owned(),
        ));
        ok_agents.push(logical.to_owned());
    }
    let mut resp = HostResponse::list(ok_agents);
    resp.queued_dags = Some(queued);
    Ok(resp)
}

/// Stop the given `agents` (resolved logical names) then `infra` containers
/// (`hivectl stop`). Agents go down before infra so they're not mid-request
/// against a forge/matrix that's already gone. Per-target failures are
/// aggregated rather than aborting on the first error, mirroring
/// `handle_restart_all`. Callers resolve the [`LifecycleScope`] to these
/// explicit name lists up front — this never sees the "all" flag.
///
/// Every agent rides the job queue: a `graceful` stop submits the
/// quiesce DAG (signal → drain → reconcile-stop; all drains overlap),
/// a hard stop a plain stop DAG — both persist `wanted = Offline` and
/// serialize on the agent's lease so nothing races an in-flight
/// rebuild. The response carries the DAG ids so `hivectl` can wait
/// with per-node progress. Infra containers have no harness / lease
/// and stay direct + synchronous.
async fn handle_stop(
    coord: &Arc<Coordinator>,
    agents: &[String],
    infra: &[InfraContainer],
    graceful: bool,
) -> Result<HostResponse> {
    tracing::info!(?agents, ?infra, graceful, "stop");
    let mut ok_items: Vec<String> = Vec::new();
    let mut errors: Vec<String> = Vec::new();
    let mut queued: Vec<u64> = Vec::new();

    for agent in agents {
        let reason = if graceful {
            "manual via hivectl graceful stop"
        } else {
            "manual via hivectl stop"
        };
        let id = if graceful {
            crate::job_queue::submit::graceful_stop(
                coord,
                agent,
                crate::job_queue::Source::Manual,
                reason.to_owned(),
            )
        } else {
            crate::job_queue::submit::stop(
                coord,
                agent,
                crate::job_queue::Source::Manual,
                reason.to_owned(),
            )
        };
        queued.push(id);
        ok_items.push(agent.clone());
    }

    // Agents go down before infra so they're not mid-request against a
    // forge/matrix that's already gone. Hard stops are quick kills —
    // await their DAGs (bounded) before touching infra. Graceful stops
    // keep the immediate return (drains take minutes and the
    // agents-then-infra race pre-existed there).
    if !graceful && !infra.is_empty() {
        await_dags(coord, &queued, std::time::Duration::from_mins(2)).await;
    }
    for &container in infra {
        let name = container.unit_name();
        match crate::priv_client::control_infra_container(container, InfraAction::Stop).await {
            Ok(()) => ok_items.push(name.to_owned()),
            Err(e) => {
                tracing::warn!(%name, error = ?e, "stop: infra stop failed");
                errors.push(format!("{name}: {e:#}"));
            }
        }
    }

    let mut resp = finish_lifecycle(ok_items, &errors);
    resp.queued_dags = Some(queued);
    Ok(resp)
}

/// Best-effort server-side wait for a set of DAGs to settle terminal,
/// bounded by `timeout` — used to preserve ordering invariants inside
/// one request (agent stops before infra stops) without trusting the
/// client to wait.
async fn await_dags(coord: &Arc<Coordinator>, ids: &[u64], timeout: std::time::Duration) {
    let deadline = std::time::Instant::now() + timeout;
    loop {
        let snap = coord.job_queue.snapshot();
        let pending = ids
            .iter()
            .any(|id| snap.iter().any(|d| d.id == *id && !d.state.is_terminal()));
        if !pending {
            return;
        }
        if std::time::Instant::now() >= deadline {
            tracing::warn!(?ids, "await_dags: timed out; proceeding");
            return;
        }
        tokio::time::sleep(std::time::Duration::from_millis(250)).await;
    }
}

/// Start the given `infra` containers then `agents` (`hivectl start`) — the
/// inverse of [`handle_stop`]. Infra comes up before agents so the agents
/// find forge/matrix/gateway ready. Per-target failures aggregated. Callers
/// resolve the [`LifecycleScope`] to these explicit name lists up front.
async fn handle_start(
    coord: &Arc<Coordinator>,
    agents: &[String],
    infra: &[InfraContainer],
) -> Result<HostResponse> {
    tracing::info!(?agents, ?infra, "start");
    let mut ok_items: Vec<String> = Vec::new();
    let mut errors: Vec<String> = Vec::new();

    for &container in infra {
        let name = container.unit_name();
        match crate::priv_client::control_infra_container(container, InfraAction::Start).await {
            Ok(()) => ok_items.push(name.to_owned()),
            Err(e) => {
                tracing::warn!(%name, error = ?e, "start: infra start failed");
                errors.push(format!("{name}: {e:#}"));
            }
        }
    }

    let mut queued: Vec<u64> = Vec::new();
    for agent in agents {
        // Through the queue: persists `wanted = Up`, upgrades a
        // stale-rev start to a full rebuild, and serializes on the
        // agent's lease. Ids ride back for hivectl's wait loop.
        queued.push(crate::job_queue::submit::start(
            coord,
            agent,
            crate::job_queue::Source::Manual,
            "manual via hivectl start".to_owned(),
        ));
        ok_items.push(agent.clone());
    }

    let mut resp = finish_lifecycle(ok_items, &errors);
    resp.queued_dags = Some(queued);
    Ok(resp)
}

/// Restart containers hive-wide (`hivectl restart`) — the DAG-based
/// sibling of [`handle_stop`]/[`handle_start`], replacing the old
/// client-side stop-then-start composition (issue tracker "dagify hivectl
/// commands"). Each targeted agent gets exactly one atomic DAG submitted
/// up front: the `Restart` template (mechanical stop + reconcile) in the
/// common case, or `GracefulRestart` (signal → drain → mechanical stop →
/// reconcile) with `graceful` set — no "submit a DAG, wait for it, submit
/// another" composition on either path, so a dropped `hivectl` connection
/// never strands an agent, the same way `handle_restart_all` already
/// avoids it. Infra containers have no lease/DAG and restart
/// synchronously (stop then start).
async fn handle_restart_scoped(
    coord: &Arc<Coordinator>,
    scope: &LifecycleScope,
    graceful: bool,
) -> Result<HostResponse> {
    tracing::info!(?scope, graceful, "restart");
    let agents = scoped_agents(scope).await?;
    let infra = scoped_infra(scope);
    let mut ok_items: Vec<String> = Vec::new();
    let mut errors: Vec<String> = Vec::new();
    let mut queued: Vec<u64> = Vec::new();

    // One atomic DAG per agent, submitted up front — `Restart`
    // (mechanical stop + reconcile) or, with `--graceful`,
    // `GracefulRestart` (signal → drain → mechanical stop → reconcile).
    // No client- or server-side "submit a stop DAG, await it, then
    // submit a start DAG" composition: that window is exactly the
    // dropped-connection gap this DAG-based path exists to close.
    for agent in &agents {
        let id = if graceful {
            crate::job_queue::submit::graceful_restart(
                coord,
                agent,
                crate::job_queue::Source::Manual,
                "manual via hivectl restart --graceful".to_owned(),
            )
        } else {
            crate::job_queue::submit::restart(
                coord,
                agent,
                crate::job_queue::Source::Manual,
                "manual restart via hivectl restart".to_owned(),
            )
        };
        queued.push(id);
        ok_items.push(agent.clone());
    }

    for &container in &infra {
        let name = container.unit_name();
        let res = async {
            crate::priv_client::control_infra_container(container, InfraAction::Stop).await?;
            crate::priv_client::control_infra_container(container, InfraAction::Start).await
        }
        .await;
        match res {
            Ok(()) => ok_items.push(name.to_owned()),
            Err(e) => {
                tracing::warn!(%name, error = ?e, "restart: infra restart failed");
                errors.push(format!("{name}: {e:#}"));
            }
        }
    }

    let mut resp = finish_lifecycle(ok_items, &errors);
    resp.queued_dags = Some(queued);
    Ok(resp)
}

/// Resolve which sub-agent logical names a scope targets: every live
/// container (from `lifecycle::list`) when `agents` is set or the scope is
/// "everything", plus any explicit `agent_names`. Returns de-duplicated
/// logical names with the `h-` container prefix stripped.
/// A scope that targets *every* agent rather than an explicit
/// `--agent <name>` list: either the `agents` flag or a bare
/// "everything" scope. Broad scopes are the ones whose stop/start pair
/// drives the previously-running restore set (see `handle` Stop/Start
/// arms); a targeted `--agent` stop/start must not redefine it.
fn is_broad_scope(scope: &LifecycleScope) -> bool {
    scope.agents || scope.is_everything()
}

/// Assemble this hive's domain + browser-facing web URLs from c0re's
/// service env (injected by hive-c0re.nix). Each field is `None` when its
/// surface isn't browser-reachable (domain unset, forge not behind the
/// gateway, matrix GUI off), so the CLI can hint precisely instead of
/// opening a dead link. Scheme matches the existing `HIVE_FORGE_PUBLIC_URL`
/// convention (gateway terminates TLS, so https).
fn hive_urls() -> hive_host_sock::HiveUrls {
    // Treat an empty env value as unset everywhere — an empty domain would
    // otherwise render `swarm.peers."" = …` (invalid nix) and `https:///`.
    let env = |k: &str| std::env::var(k).ok().filter(|v| !v.is_empty());
    let domain = env("HYPERHIVE_HIVE_DOMAIN");
    hive_host_sock::HiveUrls {
        home: domain.as_ref().map(|d| format!("https://{d}/")),
        forge: env("HIVE_FORGE_PUBLIC_URL"),
        matrix: env("HIVE_MATRIX_PUBLIC_URL"),
        domain,
    }
}

async fn scoped_agents(scope: &LifecycleScope) -> Result<Vec<String>> {
    use std::collections::BTreeSet;
    let mut set: BTreeSet<String> = BTreeSet::new();
    if is_broad_scope(scope) {
        for c in lifecycle::list().await? {
            let logical = c
                .strip_prefix(lifecycle::AGENT_PREFIX)
                .unwrap_or(&c)
                .to_owned();
            set.insert(logical);
        }
    }
    for n in &scope.agent_names {
        set.insert(n.clone());
    }
    Ok(set.into_iter().collect())
}

/// Resolve which infra containers a scope targets. An "everything" scope
/// (no flags set) selects all controllable infra; otherwise each set flag
/// maps to its [`InfraContainer`]. Fixed order for deterministic output.
fn scoped_infra(scope: &LifecycleScope) -> Vec<InfraContainer> {
    let everything = scope.is_everything();
    let mut out = Vec::new();
    if everything || scope.ci {
        out.push(InfraContainer::Ci);
    }
    if everything || scope.forge {
        out.push(InfraContainer::Forge);
    }
    if everything || scope.gateway {
        out.push(InfraContainer::Gateway);
    }
    if everything || scope.matrix {
        out.push(InfraContainer::Matrix);
    }
    out
}

/// Build the aggregated lifecycle response: `ok` with the touched names when
/// every target succeeded, otherwise `ok: false` with the joined errors and
/// the partial success list (matches `handle_restart_all`).
fn finish_lifecycle(ok_items: Vec<String>, errors: &[String]) -> HostResponse {
    if errors.is_empty() {
        HostResponse::list(ok_items)
    } else {
        HostResponse {
            ok: false,
            error: Some(errors.join("; ")),
            agents: Some(ok_items),
            ..HostResponse::default()
        }
    }
}