applied-cli 0.6.2__tar.gz → 0.6.4__tar.gz

{applied_cli-0.6.2 → applied_cli-0.6.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: applied-cli
-Version: 0.6.2
+Version: 0.6.4
 Summary: CLI and shared client library for Applied Labs AI support agents
 Author: Applied Labs
 License-Expression: MIT

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli/cli.py

@@ -1916,6 +1916,18 @@ def scenario_bulk_run(
     target_agent_id: str = typer.Option(
         None, "--target-agent-id", help="Optional target agent for reruns"
     ),
+    contact_email: str = typer.Option(
+        None,
+        "--contact-email",
+        help="Run as a contact with this email (resolves/creates it) so test "
+        "conversations carry an email — fixes 'Email is not present' failures",
+    ),
+    contact_id: str = typer.Option(
+        None, "--contact-id", help="Run scenarios as this existing contact UUID"
+    ),
+    anonymous: bool = typer.Option(
+        False, "--anonymous", help="Run with an anonymous contact"
+    ),
     shop_id: str = typer.Option(None, "--shop-id", help="Override shop ID"),
     format: str = typer.Option(
         "text", "--format", "-f", help="Output format: text or json"
@@ -1929,6 +1941,9 @@ def scenario_bulk_run(
             scenario_ids=_parse_csv_option(scenario_ids),
             benchmark_id=benchmark_id,
             target_agent_id=target_agent_id,
+            contact_email=contact_email,
+            contact_id=contact_id,
+            anonymous=anonymous,
             output_format=format,
         )
     )
@@ -1956,6 +1971,27 @@ def scenario_bulk_status(
     typer.echo(result)
 
 
+@app.command("scenario-bulk-cancel")
+def scenario_bulk_cancel(
+    job_id: str = typer.Argument(..., help="Bulk run job ID"),
+    apply: bool = typer.Option(
+        False, "--apply", help="Cancel the pending runs (default is a dry-run plan)"
+    ),
+    shop_id: str = typer.Option(None, "--shop-id", help="Override shop ID"),
+    format: str = typer.Option(
+        "text", "--format", "-f", help="Output format: text or json"
+    ),
+) -> None:
+    """Cancel a stuck bulk run by deleting its queued/running scenario runs."""
+    client = get_client(shop_id=shop_id)
+    result = asyncio.run(
+        tools.scenario_bulk_cancel(
+            client, job_id=job_id, apply=apply, output_format=format
+        )
+    )
+    typer.echo(result)
+
+
 @app.command("audit-metric-fields")
 def audit_metric_fields(
     is_active: bool = typer.Option(

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli/tools.py

@@ -6082,23 +6082,68 @@ async def scenario_run_delete(
     return f"Scenario run {run_id} deleted successfully."
 
 
+async def _resolve_contact_override(
+    client: AppliedClient,
+    *,
+    contact_override: dict | None,
+    contact_id: str | None,
+    contact_email: str | None,
+    anonymous: bool,
+) -> dict | None:
+    """Build the contact_override payload from the convenience arguments.
+
+    An explicit contact_override dict wins. Otherwise anonymous > contact_id >
+    contact_email (which resolves/creates a contact so the test conversation
+    carries a real email).
+    """
+    if contact_override:
+        return contact_override
+    if anonymous:
+        return {"mode": "anonymous"}
+    if contact_id:
+        return {"mode": "contact", "contact_id": contact_id}
+    if contact_email:
+        contact = await client.get_or_create_contact(email=contact_email)
+        resolved_id = contact.get("id")
+        if not resolved_id:
+            raise AppliedAPIError(
+                f"Could not resolve a contact for email {contact_email}.",
+                status_code=404,
+            )
+        return {"mode": "contact", "contact_id": str(resolved_id)}
+    return None
+
+
 async def scenario_bulk_run(
     client: AppliedClient,
     scenario_ids: list[str] | None = None,
     benchmark_id: str | None = None,
     target_agent_id: str | None = None,
     contact_override: dict | None = None,
+    contact_id: str | None = None,
+    contact_email: str | None = None,
+    anonymous: bool = False,
     output_format: str = "text",
 ) -> str:
     """
     Run multiple scenarios at once.
 
+    By default a scenario run reuses the input conversation's contact, which on
+    test/benchmark conversations often has no email — causing agents to respond
+    "Email is not present in the conversation." Pass contact_email or contact_id
+    to run the scenarios as a contact that has an email, so the test conversation
+    carries it.
+
     Args:
         client: Authenticated AppliedClient
         scenario_ids: List of scenario UUIDs to run
         benchmark_id: Run all scenarios in this benchmark
         target_agent_id: Optional agent to run against (for A/B testing)
-        contact_override: Optional contact override, e.g. {"mode": "contact", "contact_id": "<uuid>"}
+        contact_override: Raw override, e.g. {"mode": "contact", "contact_id": "<uuid>"}
+            (takes precedence over the convenience args below)
+        contact_id: Run scenarios as this existing contact (gives test convos its email)
+        contact_email: Resolve/create a contact with this email and run as them
+        anonymous: Run with an anonymous contact (mode='anonymous')
 
     Returns:
         Summary of runs created
@@ -6123,10 +6168,17 @@ async def scenario_bulk_run(
             )
 
     try:
+        effective_override = await _resolve_contact_override(
+            client,
+            contact_override=contact_override,
+            contact_id=contact_id,
+            contact_email=contact_email,
+            anonymous=anonymous,
+        )
         result = await client.bulk_run_scenarios(
             scenario_ids=resolved_scenario_ids,
             target_agent_id=target_agent_id,
-            contact_override=contact_override,
+            contact_override=effective_override,
         )
     except AppliedAPIError as e:
         return _format_error(e)
@@ -6230,6 +6282,95 @@ async def scenario_bulk_status(
     return output
 
 
+_PENDING_RUN_STATUSES = ("queued", "running")
+
+
+async def scenario_bulk_cancel(
+    client: AppliedClient,
+    job_id: str,
+    *,
+    apply: bool = False,
+    output_format: str = "text",
+) -> str:
+    """
+    Cancel a stuck bulk scenario run job.
+
+    A bulk run can get stuck with runs left in 'queued' or 'running' forever, so
+    the job never reports complete. This cancels the job by deleting those pending
+    runs; completed and failed runs (and their result conversations) are preserved.
+    Once no pending runs remain, the job reports as finished.
+
+    This replaces the manual workaround of deleting the 'agent-test-bulk-job:'
+    key from browser localStorage.
+
+    Args:
+        client: Authenticated AppliedClient
+        job_id: The bulk run job UUID
+        apply: If True, delete the pending runs; otherwise just report the plan
+        output_format: 'text' (default) or 'json'
+
+    Returns:
+        Summary of pending runs cancelled (or that would be cancelled).
+    """
+    try:
+        runs = await client.list_scenario_runs(
+            bulk_job_id=job_id, latest=False, limit=1000
+        )
+    except AppliedAPIError as e:
+        return _format_error(e)
+
+    pending = [r for r in runs if str(r.get("status") or "") in _PENDING_RUN_STATUSES]
+    result: dict[str, Any] = {
+        "job_id": job_id,
+        "apply": apply,
+        "total_runs": len(runs),
+        "pending_runs": len(pending),
+        "terminal_runs": len(runs) - len(pending),
+        "cancelled": 0,
+        "errors": [],
+    }
+
+    if not runs:
+        result["message"] = (
+            f"No runs found for job {job_id} — it may already be cleared or the "
+            f"job id is invalid. Nothing to cancel."
+        )
+        return to_json(result) if output_format == "json" else result["message"]
+
+    if not pending:
+        result["message"] = (
+            f"Job {job_id} has no pending (queued/running) runs — nothing to "
+            f"cancel. {result['terminal_runs']} run(s) already finished."
+        )
+        return to_json(result) if output_format == "json" else result["message"]
+
+    if not apply:
+        result["message"] = (
+            f"Would cancel {len(pending)} pending run(s) (queued/running) for job "
+            f"{job_id}; {result['terminal_runs']} finished run(s) preserved. "
+            f"Re-run with --apply to cancel."
+        )
+        return to_json(result) if output_format == "json" else result["message"]
+
+    cancelled = 0
+    for run in pending:
+        run_id = run.get("id")
+        if not run_id:
+            continue
+        try:
+            await client.delete_scenario_run(str(run_id))
+            cancelled += 1
+        except AppliedAPIError as e:
+            result["errors"].append({"run_id": str(run_id), "error": str(e)})
+    result["cancelled"] = cancelled
+    result["message"] = (
+        f"Cancelled {cancelled} pending run(s) for job {job_id}; "
+        f"{result['terminal_runs']} finished run(s) preserved."
+        + (f" {len(result['errors'])} error(s)." if result["errors"] else "")
+    )
+    return to_json(result) if output_format == "json" else result["message"]
+
+
 async def _load_audit_target_summaries(
     client: AppliedClient,
     ratings: list[dict[str, Any]],

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli/v2/domains.py

@@ -119,6 +119,7 @@ DOMAIN_TOOL_RENAMES: dict[str, dict[str, str]] = {
         "scenario_run_delete": "scenarios_runs_delete",
         "scenario_bulk_run": "scenarios_bulk_run",
         "scenario_bulk_status": "scenarios_bulk_status",
+        "scenario_bulk_cancel": "scenarios_bulk_cancel",
     },
     "taxonomy": {
         "taxonomy_list": "taxonomy_items_list",

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli/v2/scenarios.py

@@ -27,6 +27,14 @@ class ScenariosBulkRunInput(StrictInput):
     benchmark_id: str | None = None
     target_agent_id: str | None = None
     contact_override: dict[str, Any] | None = None
+    contact_id: str | None = None
+    contact_email: str | None = None
+    anonymous: bool = False
+
+
+class ScenariosBulkCancelInput(StrictInput):
+    job_id: str
+    apply: bool = False
 
 
 class BenchmarksListInput(StrictInput):
@@ -791,10 +799,19 @@ async def scenarios_bulk_run_handler(
             )
 
     try:
+        from applied_cli.tools import _resolve_contact_override
+
+        effective_override = await _resolve_contact_override(
+            client,
+            contact_override=params.contact_override,
+            contact_id=params.contact_id,
+            contact_email=params.contact_email,
+            anonymous=params.anonymous,
+        )
         result = await client.bulk_run_scenarios(
             scenario_ids=resolved_scenario_ids,
             target_agent_id=params.target_agent_id,
-            contact_override=params.contact_override,
+            contact_override=effective_override,
         )
     except AppliedAPIError as exc:
         return _api_error_result(exc)
@@ -850,6 +867,38 @@ async def scenarios_bulk_status_handler(
     )
 
 
+async def scenarios_bulk_cancel_handler(
+    client: AppliedClient,
+    params: ScenariosBulkCancelInput,
+) -> ToolResult[Any]:
+    from applied_cli import tools as legacy_tools
+
+    raw = await legacy_tools.scenario_bulk_cancel(
+        client,
+        job_id=params.job_id,
+        apply=params.apply,
+        output_format="json",
+    )
+    try:
+        data = json.loads(raw)
+    except (json.JSONDecodeError, TypeError):
+        return ToolResult(data={"message": raw}, summary=str(raw))
+
+    next_actions = []
+    if not params.apply and data.get("pending_runs"):
+        next_actions.append("Re-run with apply=true to cancel the pending runs.")
+    elif data.get("cancelled"):
+        next_actions.append(
+            f"Poll scenarios_bulk_status with job_id='{params.job_id}' to confirm "
+            f"the job now reports finished."
+        )
+    return ToolResult(
+        data=data,
+        summary=data.get("message", "Bulk cancel processed."),
+        next_actions=next_actions,
+    )
+
+
 def scenario_specs() -> list[ToolSpec]:
     return [
         ToolSpec(
@@ -1013,7 +1062,9 @@ def scenario_specs() -> list[ToolSpec]:
             namespace="scenarios",
             description=(
                 "Run selected scenarios or every scenario in a benchmark and "
-                "return the queued job metadata."
+                "return the queued job metadata. Pass contact_email or contact_id "
+                "to run as a contact with an email (fixes 'Email is not present' "
+                "failures on test conversations)."
             ),
             input_model=ScenariosBulkRunInput,
             output_model=None,
@@ -1034,4 +1085,18 @@ def scenario_specs() -> list[ToolSpec]:
             read_write_mode="read",
             tags=["scenario_bulk_status", "native"],
         ),
+        ToolSpec(
+            name="scenarios_bulk_cancel",
+            namespace="scenarios",
+            description=(
+                "Cancel a stuck bulk scenario run job by deleting its queued/"
+                "running runs (completed/failed runs are preserved). Dry-run by "
+                "default; set apply=true to cancel."
+            ),
+            input_model=ScenariosBulkCancelInput,
+            output_model=None,
+            handler=scenarios_bulk_cancel_handler,
+            read_write_mode="write",
+            tags=["scenario_bulk_cancel", "native"],
+        ),
     ]

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: applied-cli
-Version: 0.6.2
+Version: 0.6.4
 Summary: CLI and shared client library for Applied Labs AI support agents
 Author: Applied Labs
 License-Expression: MIT

{applied_cli-0.6.2 → applied_cli-0.6.4}/applied_cli.egg-info/SOURCES.txt

@@ -49,6 +49,8 @@ tests/test_conversation_tools.py
 tests/test_flow_tools.py
 tests/test_knowledge_content_tools.py
 tests/test_recovery.py
+tests/test_scenario_bulk_cancel.py
+tests/test_scenario_bulk_run_contact.py
 tests/test_toolkit_contract.py
 tests/test_v2_agents.py
 tests/test_v2_articles.py

{applied_cli-0.6.2 → applied_cli-0.6.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "applied-cli"
-version = "0.6.2"
+version = "0.6.4"
 description = "CLI and shared client library for Applied Labs AI support agents"
 readme = "README.md"
 requires-python = ">=3.11"

applied_cli-0.6.4/tests/test_scenario_bulk_cancel.py

@@ -0,0 +1,87 @@
+import json
+
+import pytest
+
+from applied_cli import tools
+
+
+class FakeBulkClient:
+    def __init__(self, runs):
+        self._runs = runs
+        self.deleted = []
+
+    async def list_scenario_runs(
+        self, scenario_id=None, benchmark_id=None, bulk_job_id=None, latest=False,
+        limit=50,
+    ):
+        return list(self._runs)
+
+    async def delete_scenario_run(self, run_id):
+        self.deleted.append(run_id)
+
+
+STUCK_RUNS = [
+    {"id": "r1", "status": "completed"},
+    {"id": "r2", "status": "queued"},
+    {"id": "r3", "status": "running"},
+    {"id": "r4", "status": "failed"},
+]
+
+
+@pytest.mark.asyncio
+async def test_dry_run_reports_pending_without_deleting():
+    client = FakeBulkClient(STUCK_RUNS)
+    result = await tools.scenario_bulk_cancel(client, "job-1", output_format="json")
+    data = json.loads(result)
+    assert data["pending_runs"] == 2
+    assert data["terminal_runs"] == 2
+    assert data["cancelled"] == 0
+    assert client.deleted == []  # dry run deletes nothing
+
+
+@pytest.mark.asyncio
+async def test_apply_deletes_only_pending_runs():
+    client = FakeBulkClient(STUCK_RUNS)
+    result = await tools.scenario_bulk_cancel(
+        client, "job-1", apply=True, output_format="json"
+    )
+    data = json.loads(result)
+    assert data["cancelled"] == 2
+    # Only the queued/running runs are deleted; completed/failed preserved.
+    assert set(client.deleted) == {"r2", "r3"}
+
+
+@pytest.mark.asyncio
+async def test_no_pending_runs_is_a_noop():
+    client = FakeBulkClient(
+        [{"id": "r1", "status": "completed"}, {"id": "r2", "status": "failed"}]
+    )
+    result = await tools.scenario_bulk_cancel(
+        client, "job-1", apply=True, output_format="text"
+    )
+    assert "nothing to" in result.lower()
+    assert client.deleted == []
+
+
+@pytest.mark.asyncio
+async def test_unknown_job_reports_no_runs():
+    client = FakeBulkClient([])
+    result = await tools.scenario_bulk_cancel(client, "missing", output_format="text")
+    assert "No runs found" in result
+    assert client.deleted == []
+
+
+@pytest.mark.asyncio
+async def test_v2_scenarios_bulk_cancel_handler_structured():
+    from applied_cli.v2.scenarios import (
+        ScenariosBulkCancelInput,
+        scenarios_bulk_cancel_handler,
+    )
+
+    client = FakeBulkClient(STUCK_RUNS)
+    result = await scenarios_bulk_cancel_handler(
+        client, ScenariosBulkCancelInput(job_id="job-1", apply=False)
+    )
+    assert result.data["pending_runs"] == 2
+    assert "apply=true" in " ".join(result.next_actions)
+    assert client.deleted == []

applied_cli-0.6.4/tests/test_scenario_bulk_run_contact.py

@@ -0,0 +1,116 @@
+import json
+
+import pytest
+
+from applied_cli import tools
+
+
+class FakeRunClient:
+    def __init__(self):
+        self.bulk_kwargs = None
+        self.get_or_create_calls = []
+
+    async def list_scenarios(self, benchmark_id=None, limit=500, **kwargs):
+        return [{"id": "s1"}, {"id": "s2"}]
+
+    async def get_or_create_contact(self, email=None, name=None, phone=None):
+        self.get_or_create_calls.append(email)
+        return {"id": "contact-123", "email": email}
+
+    async def bulk_run_scenarios(
+        self, scenario_ids=None, target_agent_id=None, contact_override=None
+    ):
+        self.bulk_kwargs = {
+            "scenario_ids": scenario_ids,
+            "target_agent_id": target_agent_id,
+            "contact_override": contact_override,
+        }
+        return {
+            "job_id": "job-1",
+            "total": len(scenario_ids or []),
+            "queued": len(scenario_ids or []),
+            "scenario_run_ids": ["r1", "r2"],
+            "contact_override": contact_override,
+        }
+
+
+@pytest.mark.asyncio
+async def test_no_contact_args_sends_no_override():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(client, benchmark_id="bench-1", output_format="json")
+    assert client.bulk_kwargs["contact_override"] is None
+
+
+@pytest.mark.asyncio
+async def test_contact_id_builds_override():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", contact_id="c-9", output_format="json"
+    )
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "c-9",
+    }
+
+
+@pytest.mark.asyncio
+async def test_contact_email_resolves_then_overrides():
+    client = FakeRunClient()
+    result = await tools.scenario_bulk_run(
+        client,
+        benchmark_id="bench-1",
+        contact_email="casey@example.com",
+        output_format="json",
+    )
+    assert client.get_or_create_calls == ["casey@example.com"]
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "contact-123",
+    }
+    assert json.loads(result)["job_id"] == "job-1"
+
+
+@pytest.mark.asyncio
+async def test_anonymous_mode():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", anonymous=True, output_format="json"
+    )
+    assert client.bulk_kwargs["contact_override"] == {"mode": "anonymous"}
+
+
+@pytest.mark.asyncio
+async def test_explicit_override_wins_over_convenience_args():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client,
+        benchmark_id="bench-1",
+        contact_override={"mode": "contact", "contact_id": "raw"},
+        contact_email="ignored@example.com",
+        output_format="json",
+    )
+    # The raw override is used; email resolution is skipped.
+    assert client.get_or_create_calls == []
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "raw",
+    }
+
+
+@pytest.mark.asyncio
+async def test_v2_handler_threads_contact_email():
+    from applied_cli.v2.scenarios import (
+        ScenariosBulkRunInput,
+        scenarios_bulk_run_handler,
+    )
+
+    client = FakeRunClient()
+    result = await scenarios_bulk_run_handler(
+        client,
+        ScenariosBulkRunInput(benchmark_id="bench-1", contact_email="x@example.com"),
+    )
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "contact-123",
+    }
+    assert result.data["job_id"] == "job-1"