shopdb-flask/plugins/endpoints/commands.py

"""Server-side allowlist of endpoint agent commands.

The WebUI never sends an arbitrary command string. It sends an allowlisted
ACTION NAME; the concrete on-endpoint behavior (which script to run, how to stop
a process) lives in the agent. This module is the single source of truth for
which action names are legal and what they mean, so both the enqueue route and
the settings/UI surface validate against the same set.

The structure anticipates per-PC-type allowlists: today every computer may run
every action in COMMAND_ALLOWLIST, but ALLOWED_ACTIONS_FOR is the seam a later
change swaps for a type-scoped lookup (inspection PCs vs shopfloor PCs).
"""

# action name -> operator-facing metadata. label/description feed the UI; the
# agent maps the action name to its local handler (see CommandExecutor).
#
# Two families of action:
#   * Named MTConnect actions - fixed target, no parameters. The original,
#     bay-specific controls kept for the inline services-panel toggle.
#   * Generic RMM actions (service_* / process_*) - the target is supplied by
#     the policy action's params ('target' = a Windows service name or a process
#     image like 'DncMain.exe'). This is what populates the Configuration Policy
#     builder with reusable controls that any scope (PC type or application) can
#     attach. The 'targetkind' hint tells the builder which target field to show.
#     NOTE: agent-side execution of the generic actions is not wired yet; the
#     server allowlist and policy plumbing land first (the whole feature is
#     flag-gated), so a generic action saved today is authored and gated but is
#     a no-op on the endpoint until the CommandExecutor handlers ship.
COMMAND_ALLOWLIST = {
    'service_start': {
        'label': 'Start service',
        'description': 'Start a Windows service by name on the endpoint.',
        'targetkind': 'service',
    },
    'service_stop': {
        'label': 'Stop service',
        'description': 'Stop a running Windows service by name on the endpoint.',
        'targetkind': 'service',
    },
    'service_status': {
        'label': 'Check service',
        'description': 'Report whether a named Windows service is running.',
        'targetkind': 'service',
    },
    'process_start': {
        'label': 'Start process',
        'description': 'Launch a process by executable path on the endpoint.',
        'targetkind': 'process',
    },
    'process_stop': {
        'label': 'Stop process',
        'description': 'Stop a running process by image name on the endpoint.',
        'targetkind': 'process',
    },
    'process_status': {
        'label': 'Check process',
        'description': 'Report whether a named process is running.',
        'targetkind': 'process',
    },
    'mtconnect_start': {
        'label': 'Start MTConnect',
        'description': 'Run the MTConnect UDC agent start script on the endpoint.',
    },
    'mtconnect_stop': {
        'label': 'Stop MTConnect',
        'description': 'Stop the running MTConnect UDC agent on the endpoint.',
    },
    'mtconnect_status': {
        'label': 'Check MTConnect',
        'description': 'Report whether the MTConnect UDC agent is running.',
    },
    'system_restart': {
        'label': 'Restart computer',
        'description': 'Restart the endpoint computer immediately.',
    },
    'collect_now': {
        'label': 'Collect now',
        'description': 'Force the endpoint agent to report fresh telemetry off-cadence.',
    },
    'network_connections': {
        'label': 'Show network connections',
        'description': 'Return the full TCP connection table with owning processes. '
                       'Read-only diagnostic; the standard report carries only '
                       'listening ports because established connections churn.',
    },
    'scheduledtask_run': {
        'label': 'Run scheduled task',
        'description': 'Start an already-registered Windows scheduled task on '
                       'the endpoint by its task path (for example '
                       '\\ReportIP). The agent does not create, modify, or '
                       'define tasks - it only starts one that already exists, '
                       'and only when the path clears the agent-side approved '
                       'task allowlist. The task runs in whatever context its '
                       'own registered definition specifies, not the agent\'s.',
        'targetkind': 'scheduledtask',
    },
    'run_script': {
        'label': 'Run script',
        'description': 'Run a ShopDB-hosted script on the endpoint. The agent '
                       'fetches the script body by digest using its collector '
                       'credential and runs it in the script\'s configured '
                       'context (system or the logged-on user). CAUTION: a '
                       'script sent this way runs as SYSTEM unless its '
                       'descriptor pins the user context, so this is arbitrary '
                       'code execution on the endpoint by design. Grant '
                       'endpoints.deploy only to the most trusted operators, '
                       'and gate the UI behind an explicit confirm the way '
                       'system_restart is gated (see confirmRestart in '
                       'PCDetail.vue).',
        'targetkind': 'script',
    },
    'update_app': {
        'label': 'Install/update application',
        'description': 'Push a ShopDB-hosted installer for an application (by app '
                       'or by a specific version). The agent fetches the pinned '
                       'package by digest using its collector credential and runs '
                       'it silently.',
        'targetkind': 'app',
    },
}


# Dell Command | Update: TWO utilities, deliberately two separate actions.
# A scan reports what is available and changes nothing; an update installs and
# can reboot. Collapsing them into one parameterized action would make the
# destructive one reachable by editing a field, so they stay distinct and the
# UI can gate them differently.
#
# Both are driven through scheduledtask_run against ShopDB-owned task
# definitions (\ShopDB-DellCommandScan and \ShopDB-DellCommandUpdate, already in
# ScheduledTaskPolicy's shipped defaults) rather than a bespoke verb. The task
# definition - not the agent - pins the dcu-cli.exe path, its arguments, and its
# output path, which is why no new agent code is needed to run them.
#
# OUTPUT PATH CONVENTION. A ShopDB-initiated scan MUST write its report under
# %ProgramData%\ShopDB\EndpointAgent\ (the same SYSTEM-owned root as the agent
# logs), NOT to Dell's default location. Corporate-managed DCU tasks write to
# the default; keeping ShopDB's output separate is the only thing that makes
# "did ShopDB ask for this scan, or did corporate" answerable when the results
# are collected. The path lives in the task definition and is generated
# server-side, so it can change without touching the agent.
#
# STILL TO DO - the results collector. Firing DCU is the easy half. schtasks
# /run returns as soon as the task STARTS, so a successful scheduledtask_run
# means "scan started", never "scan finished" - the command result cannot carry
# findings. Reporting available updates back into ShopDB needs a collector that
# reads the report from the path above and folds it into the telemetry payload.
# That is a separate change from this one; do not infer from a green command
# result that the data path exists.


# Reserved seam - one-click mapped-drive remediation. NOT implemented; this is a
# placement note so the next person adds it here rather than inventing a new
# transport. When it lands it will re-establish a Disconnected mapping surfaced
# by the user-context card:
#
#     'remap_drive': {
#         'label': 'Reconnect mapped drive',
#         'description': 'Re-establish a disconnected network drive mapping.',
#         'targetkind': 'drive',
#     },
#
# It MUST execute in the interactive USER session. A SYSTEM-context remap creates
# a mapping in SYSTEM's own drive table that the operator cannot see, so the card
# would keep reporting the drive as broken while the "fix" silently succeeded
# somewhere nobody is looking. An ELEVATED remap fails the same way, because UAC
# token separation gives the elevated context a different drive table - the same
# reason the probe itself runs unelevated from the tray.
#
# CommandExecutor.RunScriptInUserSession is the existing path for this.


# Actions that demand a permission stricter than the base endpoints.command.
# The enqueue route requires the caller to ALSO hold the mapped permission, so a
# destructive action (a remote reboot) is a heavier grant than a routine toggle.
# NOTE FOR THE ROLES/PERMISSIONS WORK: scheduledtask_run is deliberately mapped
# onto the EXISTING endpoints.deploy grant rather than introducing a new one, so
# the wire is pre-built and needs no migration. Be aware this WIDENS what
# endpoints.deploy means: it previously covered run_script + update_app, and now
# also covers starting approved scheduled tasks. If that bundle is too coarse,
# the clean split is a separate endpoints.corptools permission mapped here - no
# other code needs to change, because the enqueue route reads this table.
#
# Suggested operator-facing wording for the combined grant:
#   "Caution: Operator can run scheduled tasks, update applications, and affect
#    system-level changes to the endpoint."
# That wording is accurate for scheduledtask_run and update_app. It UNDERSTATES
# run_script, which is arbitrary SYSTEM code execution - if run_script stays in
# this bundle, say so plainly in the role description.
ELEVATED_ACTION_PERMISSIONS = {
    'system_restart': 'endpoints.restart',
    'scheduledtask_run': 'endpoints.deploy',
    'run_script': 'endpoints.deploy',
    'update_app': 'endpoints.deploy',
}


def elevated_permission_for(action):
    """The extra permission an action requires, or None for routine actions."""
    return ELEVATED_ACTION_PERMISSIONS.get(action)


# Maps a monitored services/processes row to the start/stop/status actions that
# control it, so the services panel can render an inline toggle on exactly the
# rows that have a registered command action. The key is the agent-reported
# 'app' label; 'kind' pins the toggle to the authoritative row (for MTConnect
# the agent.exe PROCESS is the source of truth - run.bat launches it and the
# Windows service state is not reliable), so the service row stays badge-only.
ROW_COMMAND_TARGETS = {
    ('MTConnect', 'process'): {
        'start': 'mtconnect_start',
        'stop': 'mtconnect_stop',
        'status': 'mtconnect_status',
    },
}


def row_actions_for(app, kind, name=None):
    """Return the {start, stop, status, target, targetkind} map for a row.

    Two paths:
      * A row registered in ROW_COMMAND_TARGETS (MTConnect) keeps its fixed,
        named actions with no wire target - agent.exe is the source of truth and
        the behavior is pinned in the agent.
      * Any other process/service row maps to the generic primitives
        (process_* / service_*) with the row's OWN reported name as the wire
        target. This is what lets a new app light up start/stop/status without
        an agent rebuild: the target rides the wire.

    Rows with no usable kind return None and render as a plain status badge.
    """
    if not app and not name:
        return None
    fixed = ROW_COMMAND_TARGETS.get((app, (kind or '').lower()))
    if fixed:
        return {**fixed, 'target': None, 'targetkind': None}
    normalized = (kind or '').lower()
    if normalized == 'process' and name:
        return {
            'start': 'process_start',
            'stop': 'process_stop',
            'status': 'process_status',
            'target': name,
            'targetkind': 'process',
        }
    if normalized == 'service' and name:
        return {
            'start': 'service_start',
            'stop': 'service_stop',
            'status': 'service_status',
            'target': name,
            'targetkind': 'service',
        }
    return None


def is_allowed_action(action):
    """True when action is a known, allowlisted command name."""
    return action in COMMAND_ALLOWLIST


def action_targetkind(action):
    """The targetkind an action requires ('service'/'process'), or None.

    Generic actions declare a targetkind in COMMAND_ALLOWLIST; fixed-target
    actions (mtconnect_*) declare none and must not carry a wire target.
    """
    meta = COMMAND_ALLOWLIST.get(action) or {}
    return meta.get('targetkind')


def requires_target(action):
    """True when the action is generic and needs a wire target."""
    return action_targetkind(action) is not None


# Monitor attachment as a policy action. A monitor definition is bound to a PC
# type by carrying a ConfigurationPolicyAction whose name is 'monitor:<id>'.
# Distinct names let one policy attach several monitors and fold naturally
# (first-match-wins per name), and they are validated in the policy layer but
# never enqueued as commands because ALLOWED_ACTIONS_FOR intersects with
# COMMAND_ALLOWLIST, which carries no monitor: entries.
MONITOR_ACTION_PREFIX = 'monitor:'


def monitor_action_name(endpointmonitorid):
    """Build the policy action name that attaches a monitor definition."""
    return f'{MONITOR_ACTION_PREFIX}{endpointmonitorid}'


def is_monitor_action(action):
    """True when action names a monitor attachment (monitor:<id>)."""
    return isinstance(action, str) and action.startswith(MONITOR_ACTION_PREFIX)


def monitor_id_from_action(action):
    """The monitor id an action attaches, or None when it is not a monitor action."""
    if not is_monitor_action(action):
        return None
    raw = action[len(MONITOR_ACTION_PREFIX):]
    try:
        return int(raw)
    except (TypeError, ValueError):
        return None


# Actions gated by ROLE ONLY, never by per-endpoint Configuration Policy.
#
# These are the Corporate Tools actions. They start already-registered,
# agent-allowlisted tasks and are not a per-machine configuration concern, so a
# policy that never mentions them must not remove them. Policy exists to shape
# what a given PC TYPE exposes (which monitors, which service primitives); it is
# not the authority on whether an admin may run a corporate tool.
#
# Without this exemption the intersection at the end of ALLOWED_ACTIONS_FOR
# silently drops these actions on any endpoint whose effective policy is
# non-empty, producing "Action not permitted by policy for this endpoint" even
# for a caller holding endpoints.deploy. The real gate for these stays
# ELEVATED_ACTION_PERMISSIONS (role), enforced in the enqueue route.
POLICY_EXEMPT_ACTIONS = frozenset({
    'scheduledtask_run',
})


def ALLOWED_ACTIONS_FOR(pctype=None, assetid=None):
    """The action names an endpoint may run.

    Base allowlist is flat (every action for every computer). When the computers
    Configuration/Business Policy feature is enabled AND an assetid is given, the
    result is narrowed to the actions the endpoint's effective policy permits -
    this is the allowlist-gating half of the policy design. With the feature off
    (or no assetid), it returns the full base allowlist, so nothing changes until
    a site opts in. Actions in POLICY_EXEMPT_ACTIONS are never narrowed away;
    they are gated by role alone. Policy resolution is imported lazily to keep
    this module free of a hard dependency on the computers plugin.
    """
    base = set(COMMAND_ALLOWLIST.keys())
    if assetid is None:
        return base
    try:
        from plugins.computers.services.policyresolver import effective_policy_for
    except Exception:
        return base
    effective = effective_policy_for(assetid)
    allowed = effective.allowed_actions()
    if not allowed:
        # No policy spoke for this endpoint. With the feature off the resolver
        # returns empty and we fall back to the base allowlist (unchanged
        # behavior). This is intentionally permissive; deny-by-default would be
        # a separate, explicit setting.
        return base
    # Attaching a monitor expresses intent to watch a process/service AND to
    # control it: the panel renders that row's Start/Stop/Status buttons, so the
    # gate must permit the matching generic primitives or every click 409s
    # "not permitted by policy". Expand each monitor:<id> to its kind's control
    # actions before intersecting.
    allowed = allowed | _monitor_control_actions(allowed)
    # Role-gated actions are added back unconditionally: policy narrows the
    # per-endpoint surface, but it does not own the corporate tools.
    allowed = allowed | POLICY_EXEMPT_ACTIONS
    # Never widen past the code-level allowlist: policy can only gate within it.
    return base & allowed


def _monitor_control_actions(action_names):
    """Control primitives granted by the monitor: attachments in a policy.

    A process monitor grants process_start/stop/status; a service monitor grants
    service_start/stop/status. The monitor's kind is read from the catalog, so a
    definition's kind is the single source of truth for what its row can do.
    """
    ids = set()
    for name in action_names:
        if isinstance(name, str) and name.startswith(MONITOR_ACTION_PREFIX):
            monitor_id = monitor_id_from_action(name)
            if monitor_id is not None:
                ids.add(monitor_id)
    if not ids:
        return set()

    try:
        from .models import EndpointMonitor
    except Exception:
        return set()

    granted = set()
    for monitor in EndpointMonitor.query.filter(
            EndpointMonitor.endpointmonitorid.in_(ids)).all():
        kind = (monitor.kind or '').lower()
        if kind == 'process':
            granted |= {'process_start', 'process_stop', 'process_status'}
        elif kind == 'service':
            granted |= {'service_start', 'service_stop', 'service_status'}
    return granted