Skip to content

OpenAI's Codex Security CLI and TypeScript SDK for finding, validating, and fixing security vulnerabilities. npm: https://www.npmjs.com/package/@openai/codex-security

test(workbench): share setup for scan history and usage tests (#1409)

mldangelo-oaia31c5b32b4c2
diff --git a/plugins/codex-security/tests/test_deep_security_scan_concurrency_guard.py b/plugins/codex-security/tests/test_deep_security_scan_concurrency_guard.py
index 01f0eac80..d8bdbc506 100644
--- a/plugins/codex-security/tests/test_deep_security_scan_concurrency_guard.py
+++ b/plugins/codex-security/tests/test_deep_security_scan_concurrency_guard.py
@@ -3,7 +3,12 @@ from __future__ import annotations
import sqlite3
from pathlib import Path
-from workbench_test_support import create_saved_git_workspace, run_workbench
+from workbench_test_support import (
+ create_saved_git_workspace,
+ fail_scan,
+ get_scan,
+ start_scan_command,
+)
def test_scan_context_reports_only_other_running_deep_scans(tmp_path: Path) -> None:
@@ -23,12 +28,7 @@ def test_scan_context_reports_only_other_running_deep_scans(tmp_path: Path) -> N
for name, target in targets.items()
}
scans = {
- name: run_workbench(
- state_dir,
- "start-scan",
- "--workspace-id",
- str(workspace["id"]),
- )
+ name: start_scan_command(state_dir, str(workspace["id"]))
for name, workspace in workspaces.items()
}
@@ -39,14 +39,7 @@ def test_scan_context_reports_only_other_running_deep_scans(tmp_path: Path) -> N
"UPDATE scans SET handoff_status = 'delivered' WHERE id IN (?, ?)",
(failed_scan_id, other_scan_id),
)
- run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- failed_scan_id,
- "--message",
- "Stopped for the fixture.",
- )
+ fail_scan(state_dir, failed_scan_id, "Stopped for the fixture.")
complete_scan_id = str(scans["complete"]["results"]["scanId"])
with sqlite3.connect(state_dir / "workbench.sqlite3") as connection:
connection.execute(
@@ -59,7 +52,7 @@ def test_scan_context_reports_only_other_running_deep_scans(tmp_path: Path) -> N
)
current_scan_id = str(scans["current"]["results"]["scanId"])
- context = run_workbench(state_dir, "get-scan", "--scan-id", current_scan_id)
+ context = get_scan(state_dir, current_scan_id)
assert context["otherRunningDeepScans"] == [
{
@@ -71,13 +64,6 @@ def test_scan_context_reports_only_other_running_deep_scans(tmp_path: Path) -> N
}
]
- run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- other_scan_id,
- "--message",
- "Stopped for the fixture.",
- )
- context = run_workbench(state_dir, "get-scan", "--scan-id", current_scan_id)
+ fail_scan(state_dir, other_scan_id, "Stopped for the fixture.")
+ context = get_scan(state_dir, current_scan_id)
assert context["otherRunningDeepScans"] == []
diff --git a/plugins/codex-security/tests/test_workbench_prompt_only_scan.py b/plugins/codex-security/tests/test_workbench_prompt_only_scan.py
index ea73d0a93..b7675d16a 100644
--- a/plugins/codex-security/tests/test_workbench_prompt_only_scan.py
+++ b/plugins/codex-security/tests/test_workbench_prompt_only_scan.py
@@ -17,6 +17,7 @@ from test_workbench_db import (
initialize_git_repository,
run_workbench,
)
+from workbench_test_support import get_scan, scan_command, start_scan_command, update_progress
def start_prompt_only_scan(
@@ -26,8 +27,6 @@ def start_prompt_only_scan(
*,
thread_id: str = "thread-prompt-only-scan",
mode: str = "standard",
- target_summary: str = "Prompt-only scan",
- user_context: str = "Inspect authentication boundaries",
extra_args: tuple[str, ...] = (),
) -> dict[str, object]:
return run_workbench(
@@ -42,9 +41,9 @@ def start_prompt_only_scan(
"--mode",
mode,
"--target-summary",
- target_summary,
+ "Prompt-only scan",
"--user-context",
- user_context,
+ "Inspect authentication boundaries",
"--scan-root",
str(scan_root),
*extra_args,
@@ -57,8 +56,6 @@ def start_headless_standard_scan(
scan_root: Path,
*,
thread_id: str = "thread-headless-standard-scan",
- scope: str = ".",
- target_summary: str = "Headless standard scan",
user_context: str = "Inspect authentication boundaries",
) -> dict[str, object]:
return run_workbench(
@@ -69,9 +66,9 @@ def start_headless_standard_scan(
"--target-path",
str(target),
"--scope",
- scope,
+ ".",
"--target-summary",
- target_summary,
+ "Headless standard scan",
"--user-context",
user_context,
"--scan-root",
@@ -202,10 +199,9 @@ def test_prompt_only_standard_phase_uses_latest_persisted_scan_context(
started = start_prompt_only_scan(state_dir, target, tmp_path / "scans")
scan_id = str(started["scan"]["scanId"])
updated_context = "Prioritize password-reset token validation."
- updated = run_workbench(
+ updated = scan_command(
state_dir,
"update-scan-context",
- "--scan-id",
scan_id,
"--thread-id",
"thread-prompt-only-scan",
@@ -214,14 +210,7 @@ def test_prompt_only_standard_phase_uses_latest_persisted_scan_context(
)
assert updated["scan"]["userContext"] == updated_context
- next_phase = run_workbench(
- state_dir,
- "update-progress",
- "--scan-id",
- scan_id,
- "--phase",
- "discovery",
- )
+ next_phase = update_progress(state_dir, scan_id, "--phase", "discovery")
assert next_phase["scan"]["progress"]["phase"] == "discovery"
assert next_phase["scan"]["userContext"] == updated_context
@@ -293,14 +282,7 @@ def test_prompt_only_scan_does_not_join_setup_owned_scans(tmp_path: Path) -> Non
target,
thread_id="thread-prompt-only-scan",
)
- pending = run_workbench(
- state_dir,
- "start-scan",
- "--workspace-id",
- str(saved["id"]),
- "--scan-root",
- str(scan_root),
- )
+ pending = start_scan_command(state_dir, str(saved["id"]), "--scan-root", str(scan_root))
assert pending["results"]["handoffStatus"] == "pending"
prompt_only = start_prompt_only_scan(state_dir, target, scan_root)
assert prompt_only["startDisposition"] == "created"
@@ -382,7 +364,7 @@ def test_prompt_registration_keeps_existing_scans_readable(
connection.close()
def read_scans():
- read = run_workbench(state_dir, "get-scan", "--scan-id", str(scan_id))
+ read = get_scan(state_dir, str(scan_id))
listed = run_workbench(state_dir, "list-scans")
assert read["scan"]["scanId"] == scan_id
assert any(scan["scanId"] == scan_id for scan in listed["scans"])
diff --git a/plugins/codex-security/tests/test_workbench_scan_archive.py b/plugins/codex-security/tests/test_workbench_scan_archive.py
index f7361bb05..b8066e772 100644
--- a/plugins/codex-security/tests/test_workbench_scan_archive.py
+++ b/plugins/codex-security/tests/test_workbench_scan_archive.py
@@ -4,7 +4,7 @@ import json
import sqlite3
from pathlib import Path
-from workbench_test_support import run_workbench
+from workbench_test_support import fail_scan, run_workbench
def register_scan(
@@ -39,14 +39,7 @@ def test_archived_output_preserves_old_scan_and_allows_fresh_registration(
repository.mkdir()
scan_dir.mkdir(mode=0o700)
previous = register_scan(state_dir, repository, scan_dir)
- run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- str(previous["scanId"]),
- "--message",
- "The previous scan was interrupted.",
- )
+ fail_scan(state_dir, str(previous["scanId"]), "The previous scan was interrupted.")
(scan_dir / "previous.txt").write_text("keep the previous scan\n")
report_path = scan_dir / "report.md"
report_path.write_text("# Previous scan\n")
@@ -98,13 +91,8 @@ def test_empty_failed_scan_can_be_archived_and_reused(tmp_path: Path) -> None:
repository.mkdir()
scan_dir.mkdir(mode=0o700)
previous = register_scan(state_dir, repository, scan_dir)
- run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- str(previous["scanId"]),
- "--message",
- "The previous scan failed before writing artifacts.",
+ fail_scan(
+ state_dir, str(previous["scanId"]), "The previous scan failed before writing artifacts."
)
current = register_scan(state_dir, repository, scan_dir, "--archive-existing")
@@ -129,14 +117,7 @@ def test_archive_requires_previous_directory_when_artifacts_exist(tmp_path: Path
repository.mkdir()
scan_dir.mkdir(mode=0o700)
previous = register_scan(state_dir, repository, scan_dir)
- run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- str(previous["scanId"]),
- "--message",
- "The previous scan was interrupted.",
- )
+ fail_scan(state_dir, str(previous["scanId"]), "The previous scan was interrupted.")
with sqlite3.connect(state_dir / "workbench.sqlite3") as connection:
connection.execute(
"INSERT INTO scan_artifacts (scan_id, kind, path, created_at) VALUES (?, ?, ?, ?)",
diff --git a/plugins/codex-security/tests/test_workbench_scan_history.py b/plugins/codex-security/tests/test_workbench_scan_history.py
index 6ff29fddb..f99f5ae14 100644
--- a/plugins/codex-security/tests/test_workbench_scan_history.py
+++ b/plugins/codex-security/tests/test_workbench_scan_history.py
@@ -15,9 +15,14 @@ from test_workbench_db import HEAD_CHANGED_WARNING
from test_workbench_deep_scan import begin_target_scan
from test_workbench_prompt_only_scan import start_headless_standard_scan, start_prompt_only_scan
from workbench_test_support import (
+ fail_scan,
+ get_scan,
initialize_git_repository,
mark_deep_coordinator_succeeded,
+ resume_deep_scan,
run_workbench,
+ scan_command,
+ set_triage,
stable_target_id,
write_completed_contract,
)
@@ -115,14 +120,7 @@ def create_cli_scan(
if not complete:
return launched
if mode == "deep":
- run_workbench(
- state_dir,
- "begin-deep-scan",
- "--scan-id",
- launched["scanId"],
- "--thread-id",
- "thread-scan-history",
- )
+ resume_deep_scan(state_dir, launched["scanId"], "thread-scan-history")
mark_deep_coordinator_succeeded(state_dir, launched["scanId"], scan_dir)
coverage_mode = (
@@ -175,50 +173,6 @@ def create_cli_scan(
return launched
-def insert_scan(
- connection: sqlite3.Connection,
- *,
- workspace_id: str,
- scan_id: str,
- mode: str,
- status: str,
- phase: str,
- timestamp: str,
- seal: str | None = None,
- failure: str | None = None,
- canceled: bool = False,
-) -> None:
- connection.execute(
- """
- INSERT INTO scans (
- id, workspace_id, target_path, target_revision, scope, mode,
- scan_dir, status, phase, handoff_status, failure_message,
- started_at, completed_at, created_at, updated_at, canceled_at,
- seal_manifest_digest
- ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
- """,
- (
- scan_id,
- workspace_id,
- "/tmp/target",
- "fixture-revision",
- ".",
- mode,
- f"/tmp/scans/{scan_id}",
- status,
- phase,
- "delivered",
- failure,
- timestamp,
- timestamp if status != "running" else None,
- timestamp,
- timestamp,
- timestamp if canceled else None,
- seal,
- ),
- )
-
-
def test_cli_scan_lifecycle_persists_recipes_lineage_and_filtered_history(tmp_path: Path) -> None:
state_dir = tmp_path / "state"
repository = tmp_path / "repository"
@@ -234,21 +188,22 @@ def test_cli_scan_lifecycle_persists_recipes_lineage_and_filtered_history(tmp_pa
paths=["src", "tests"],
)
failed = create_cli_scan(state_dir, root, repository, complete=False)
- run_workbench(state_dir, "fail-scan", "--scan-id", failed["scanId"], "--message", "interrupted")
+ fail_scan(state_dir, failed["scanId"], "interrupted")
assert str(uuid.UUID(first["scanId"])) == first["scanId"]
assert first["targetId"] == stable_target_id(repository)
assert first["scanDir"].startswith(str(root))
- detail = run_workbench(state_dir, "get-scan", "--scan-id", rerun["scanId"])
+ detail = get_scan(state_dir, rerun["scanId"])
assert detail["parentScanId"] == first["scanId"]
assert detail["recipe"]["target"]["paths"] == ["src", "tests"]
with sqlite3.connect(state_dir / "workbench.sqlite3") as connection:
assert connection.execute(
"SELECT parent_scan_id FROM scans WHERE id = ?", (rerun["scanId"],)
).fetchone() == (first["scanId"],)
- assert run_workbench(state_dir, "get-scan-recipe", "--scan-id", first["scanId"])["recipe"][
- "config"
- ] == {"model": "gpt-5.6-sol", "model_reasoning_effort": "high"}
+ assert scan_command(state_dir, "get-scan-recipe", first["scanId"])["recipe"]["config"] == {
+ "model": "gpt-5.6-sol",
+ "model_reasoning_effort": "high",
+ }
other = tmp_path / "other"
other.mkdir()
@@ -271,21 +226,14 @@ def test_cli_scan_persists_its_continuation_thread(tmp_path: Path) -> None:
repository = tmp_path / "repository"
repository.mkdir()
scan = create_cli_scan(state_dir, tmp_path / "results", repository, complete=False)
- initial = run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]
+ initial = get_scan(state_dir, scan["scanId"])["scan"]
assert initial["threadIds"] == []
assert initial["executionThreadIds"] == []
- result = run_workbench(
- state_dir,
- "set-scan-thread",
- "--scan-id",
- scan["scanId"],
- "--thread-id",
- "thread-1",
- )
+ result = scan_command(state_dir, "set-scan-thread", scan["scanId"], "--thread-id", "thread-1")
assert result == {"scanId": scan["scanId"], "threadId": "thread-1"}
- detail = run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])
+ detail = get_scan(state_dir, scan["scanId"])
assert detail["scan"]["continuationThreadId"] == "thread-1"
assert detail["scan"]["threadIds"] == ["thread-1"]
assert detail["scan"]["executionThreadIds"] == ["thread-1"]
@@ -299,7 +247,7 @@ def test_get_scan_keeps_desktop_standard_owners_out_of_execution_roots(tmp_path:
started = start_scan(
state_dir, repository, tmp_path / "results", thread_id="desktop-owner"
)["scan"]
- detail = run_workbench(state_dir, "get-scan", "--scan-id", started["scanId"])["scan"]
+ detail = get_scan(state_dir, started["scanId"])["scan"]
assert detail["threadIds"] == ["desktop-owner"]
assert detail["executionThreadIds"] == []
if start_scan is start_headless_standard_scan:
@@ -357,7 +305,7 @@ def test_get_scan_includes_desktop_deep_worker_threads_without_continuation(tmp_
],
)
- detail = run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]
+ detail = get_scan(state_dir, scan["scanId"])["scan"]
assert detail["continuationThreadId"] is None
assert detail["threadIds"] == [
"desktop-owner",
@@ -382,7 +330,7 @@ def test_get_scan_includes_desktop_deep_worker_threads_without_continuation(tmp_
"UPDATE scans SET continuation_thread_id = 'desktop-headless' WHERE id = ?",
(scan["scanId"],),
)
- continued = run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]
+ continued = get_scan(state_dir, scan["scanId"])["scan"]
assert continued["threadIds"] == ["desktop-headless", *detail["threadIds"]]
assert continued["executionThreadIds"] == detail["executionThreadIds"]
@@ -418,7 +366,7 @@ def test_cli_scan_preserves_original_revision_when_head_moves(tmp_path: Path) ->
subprocess.run(["git", "-C", str(repository), "add", "README.md"], check=True)
subprocess.run(["git", "-C", str(repository), "commit", "-qm", "Move HEAD"], check=True)
- completed = run_workbench(state_dir, "complete-scan", "--scan-id", launched["scanId"])
+ completed = scan_command(state_dir, "complete-scan", launched["scanId"])
assert completed["scan"]["progress"]["status"] == "complete"
assert completed["scan"]["targetRevision"] == revision
@@ -468,7 +416,7 @@ def test_cli_scan_history_persists_per_scan_cost(tmp_path: Path, context_reporti
listed = run_workbench(state_dir, "list-scans", "--repository", str(repository))
assert listed["scans"][0]["cost"] == cost
- assert run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["cost"] == cost
+ assert get_scan(state_dir, scan["scanId"])["scan"]["cost"] == cost
with sqlite3.connect(state_dir / "workbench.sqlite3") as connection:
stored = connection.execute(
"SELECT cost_json FROM scans WHERE id = ?", (scan["scanId"],)
@@ -509,18 +457,13 @@ def test_cli_scan_completion_persists_authoritative_cost_after_plugin_completion
),
)
- completed = run_workbench(
- state_dir,
- "complete-scan",
- "--scan-id",
- scan["scanId"],
- "--cost-json",
- json.dumps(cost),
+ completed = scan_command(
+ state_dir, "complete-scan", scan["scanId"], "--cost-json", json.dumps(cost)
)
assert completed["scan"]["cost"] == cost
assert completed["scan"]["usage"]["inputTokens"] == 5_000
- assert run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["cost"] == cost
+ assert get_scan(state_dir, scan["scanId"])["scan"]["cost"] == cost
def test_failed_cli_scan_history_persists_measured_cost(tmp_path: Path) -> None:
@@ -537,12 +480,9 @@ def test_failed_cli_scan_history_persists_measured_cost(tmp_path: Path) -> None:
}
scan = create_cli_scan(state_dir, tmp_path / "results", repository, complete=False)
- failed = run_workbench(
+ failed = fail_scan(
state_dir,
- "fail-scan",
- "--scan-id",
scan["scanId"],
- "--message",
"Scan stopped: cost limit exceeded.",
"--cost-json",
json.dumps(cost),
@@ -551,7 +491,7 @@ def test_failed_cli_scan_history_persists_measured_cost(tmp_path: Path) -> None:
assert failed["scan"]["cost"] == cost
assert failed["scan"]["progress"]["status"] == "failed"
assert run_workbench(state_dir, "list-scans")["scans"][0]["cost"] == cost
- assert run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["cost"] == cost
+ assert get_scan(state_dir, scan["scanId"])["scan"]["cost"] == cost
def test_scan_failure_rejects_invalid_cost_without_stopping_the_scan(tmp_path: Path) -> None:
@@ -560,26 +500,13 @@ def test_scan_failure_rejects_invalid_cost_without_stopping_the_scan(tmp_path: P
repository.mkdir()
scan = create_cli_scan(state_dir, tmp_path / "results", repository, complete=False)
- rejected = run_workbench(
- state_dir,
- "fail-scan",
- "--scan-id",
- scan["scanId"],
- "--message",
- "Scan stopped.",
- "--cost-json",
- "{}",
- check=False,
+ rejected = fail_scan(
+ state_dir, scan["scanId"], "Scan stopped.", "--cost-json", "{}", check=False
)
assert rejected["returncode"] != 0
assert "Scan cost" in rejected["stderr"]
- assert (
- run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["progress"][
- "status"
- ]
- == "running"
- )
+ assert get_scan(state_dir, scan["scanId"])["scan"]["progress"]["status"] == "running"
def test_scan_completion_rejects_invalid_cost_without_overwriting_history(tmp_path: Path) -> None:
@@ -605,19 +532,13 @@ def test_scan_completion_rejects_invalid_cost_without_overwriting_history(tmp_pa
]
for value in invalid:
- rejected = run_workbench(
- state_dir,
- "complete-scan",
- "--scan-id",
- scan["scanId"],
- "--cost-json",
- value,
- check=False,
+ rejected = scan_command(
+ state_dir, "complete-scan", scan["scanId"], "--cost-json", value, check=False
)
assert rejected["returncode"] != 0
assert "Scan cost" in rejected["stderr"]
- assert "cost" not in run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]
+ assert "cost" not in get_scan(state_dir, scan["scanId"])["scan"]
def test_scan_history_resolves_unique_prefixes_and_rejects_ambiguity(tmp_path: Path) -> None:
@@ -630,13 +551,8 @@ def test_scan_history_resolves_unique_prefixes_and_rejects_ambiguity(tmp_path: P
prefix = scan["scanId"][:8]
after_prefix = after["scanId"][:8]
- assert (
- run_workbench(state_dir, "get-scan", "--scan-id", prefix)["scan"]["scanId"]
- == scan["scanId"]
- )
- assert (
- run_workbench(state_dir, "get-scan-recipe", "--scan-id", prefix)["scanId"] == scan["scanId"]
- )
+ assert get_scan(state_dir, prefix)["scan"]["scanId"] == scan["scanId"]
+ assert scan_command(state_dir, "get-scan-recipe", prefix)["scanId"] == scan["scanId"]
compared = run_workbench(
state_dir,
"compare-scans",
@@ -669,14 +585,36 @@ def test_scan_history_resolves_unique_prefixes_and_rejects_ambiguity(tmp_path: P
workspace_id = connection.execute(
"SELECT workspace_id FROM scans WHERE id = ?", (scan["scanId"],)
).fetchone()[0]
- insert_scan(
- connection,
- workspace_id=workspace_id,
- scan_id=f"{prefix}-ffff-4000-8000-000000000000",
- mode="standard",
- status="complete",
- phase="reporting",
- timestamp="2026-07-24T00:00:00Z",
+ ambiguous_scan_id = f"{prefix}-ffff-4000-8000-000000000000"
+ timestamp = "2026-07-24T00:00:00Z"
+ connection.execute(
+ """
+ INSERT INTO scans (
+ id, workspace_id, target_path, target_revision, scope, mode,
+ scan_dir, status, phase, handoff_status, failure_message,
+ started_at, completed_at, created_at, updated_at, canceled_at,
+ seal_manifest_digest
+ ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+ """,
+ (
+ ambiguous_scan_id,
+ workspace_id,
+ "/tmp/target",
+ "fixture-revision",
+ ".",
+ "standard",
+ f"/tmp/scans/{ambiguous_scan_id}",
+ "complete",
+ "reporting",
+ "delivered",
+ None,
+ timestamp,
+ timestamp,
+ timestamp,
+ timestamp,
+ None,
+ None,
+ ),
)
for arguments in (
@@ -702,10 +640,7 @@ def test_scan_history_resolves_unique_prefixes_and_rejects_ambiguity(tmp_path: P
ambiguous = run_workbench(state_dir, *arguments, check=False)
assert ambiguous["returncode"] != 0
assert "matches multiple scans; use a longer prefix" in ambiguous["stderr"]
- assert (
- run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["scanId"]
- == scan["scanId"]
- )
+ assert get_scan(state_dir, scan["scanId"])["scan"]["scanId"] == scan["scanId"]
def test_scan_list_includes_related_git_worktrees_and_clones(tmp_path: Path) -> None:
@@ -795,16 +730,7 @@ def test_cli_scan_comparison_tracks_stable_findings_without_copying_triage(tmp_p
assert persisted["comparable"] is True
assert persisted["summary"]["persisting"] == 1
occurrence = persisted["findings"][0]["beforeOccurrenceId"]
- run_workbench(
- state_dir,
- "set-finding-triage",
- "--occurrence-id",
- occurrence,
- "--status",
- "closed",
- "--close-reason",
- "already_fixed",
- )
+ set_triage(state_dir, occurrence, "closed", "--close-reason", "already_fixed")
reopened = compare_scan_pair(state_dir, before, after)
assert reopened["summary"]["reopened"] == 1
assert reopened["findings"][0]["triage"] == {"closeReason": None, "status": "open"}
@@ -834,8 +760,7 @@ def test_scan_comparison_requires_saved_matches_and_remains_read_only(tmp_path:
known_since = connection.execute("SELECT MIN(started_at) FROM scans").fetchone()[0]
before_finding, after_finding = (
- run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"]["findings"][0]
- for scan in (before, after)
+ get_scan(state_dir, scan["scanId"])["scan"]["findings"][0] for scan in (before, after)
)
assert before_finding["findingId"] == after_finding["findingId"]
for finding, other in ((before_finding, after_finding), (after_finding, before_finding)):
@@ -948,9 +873,7 @@ def test_semantic_scan_comparison_caches_matches_and_exposes_related_findings(
assert baseline["matchingCached"] is False
assert baseline["summary"]["new"] == 1
assert baseline["summary"]["resolved"] == 1
- unmatched = run_workbench(state_dir, "get-scan", "--scan-id", before["scanId"])["scan"][
- "findings"
- ][0]
+ unmatched = get_scan(state_dir, before["scanId"])["scan"]["findings"][0]
assert "knownSince" not in unmatched
assert "knownScanIds" not in unmatched
previous = baseline["matchingInputs"]["before"][0]
@@ -990,12 +913,8 @@ def test_semantic_scan_comparison_caches_matches_and_exposes_related_findings(
assert cached["matchingCached"] is True
assert cached["matchingInputs"]["before"][0]["occurrenceId"] == previous["occurrenceId"]
- before_finding = run_workbench(state_dir, "get-scan", "--scan-id", before["scanId"])["scan"][
- "findings"
- ][0]
- after_finding = run_workbench(state_dir, "get-scan", "--scan-id", after["scanId"])["scan"][
- "findings"
- ][0]
+ before_finding = get_scan(state_dir, before["scanId"])["scan"]["findings"][0]
+ after_finding = get_scan(state_dir, after["scanId"])["scan"]["findings"][0]
assert before_finding["matches"] == [
{
"findingId": current["findingId"],
@@ -1027,9 +946,7 @@ def test_semantic_scan_comparison_caches_matches_and_exposes_related_findings(
confirmed_match(current["occurrenceId"], latest_finding["occurrenceId"]),
)
for scan in (before, after, latest):
- finding = run_workbench(state_dir, "get-scan", "--scan-id", scan["scanId"])["scan"][
- "findings"
- ][0]
+ finding = get_scan(state_dir, scan["scanId"])["scan"]["findings"][0]
assert finding["knownScanIds"] == [before["scanId"], latest["scanId"]]
assert finding["knownSince"] == known_since
@@ -1039,9 +956,7 @@ def test_semantic_scan_comparison_caches_matches_and_exposes_related_findings(
latest,
confirmed_match(previous["occurrenceId"], latest_finding["occurrenceId"]),
)
- latest_finding = run_workbench(state_dir, "get-scan", "--scan-id", latest["scanId"])["scan"][
- "findings"
- ][0]
+ latest_finding = get_scan(state_dir, latest["scanId"])["scan"]["findings"][0]
assert latest_finding["knownScanIds"] == [before["scanId"], latest["scanId"]]
assert latest_finding["knownSince"] == known_since
@@ -1100,12 +1015,9 @@ def test_semantic_scan_comparison_supports_one_to_many_without_copying_triage(
"UPDATE finding_occurrences SET severity = ? WHERE id = ?",
(("low", current[0]["occurrenceId"]), ("critical", current[1]["occurrenceId"])),
)
- run_workbench(
+ set_triage(
state_dir,
- "set-finding-triage",
- "--occurrence-id",
previous["occurrenceId"],
- "--status",
"closed",
"--close-reason",
"false_positive",
@@ -1131,13 +1043,11 @@ def test_semantic_scan_comparison_supports_one_to_many_without_copying_triage(
assert set(saved["findings"][0]["afterOccurrenceIds"]) == {
finding["occurrenceId"] for finding in current
}
- prior = run_workbench(state_dir, "get-scan", "--scan-id", before["scanId"])["scan"]["findings"][
- 0
- ]
+ prior = get_scan(state_dir, before["scanId"])["scan"]["findings"][0]
assert len(prior["matches"]) == 2
assert prior["triage"]["closeReason"] == "false_positive"
assert prior["triage"]["status"] == "closed"
- later = run_workbench(state_dir, "get-scan", "--scan-id", after["scanId"])["scan"]["findings"]
+ later = get_scan(state_dir, after["scanId"])["scan"]["findings"]
assert all(finding["triage"]["status"] == "open" for finding in later)
@@ -1173,9 +1083,7 @@ def test_uncertain_semantic_scan_matches_stay_separate(tmp_path: Path) -> None:
assert compared["summary"]["persisting"] == 0
assert all(finding["status"] == "unknown" for finding in compared["findings"])
assert all("independently reachable" in finding["reason"] for finding in compared["findings"])
- shown = run_workbench(state_dir, "get-scan", "--scan-id", after["scanId"])["scan"]["findings"][
- 0
- ]
+ shown = get_scan(state_dir, after["scanId"])["scan"]["findings"][0]
assert "matches" not in shown
assert "knownSince" not in shown
assert "knownScanIds" not in shown
@@ -1206,9 +1114,7 @@ def test_semantic_scan_comparison_replaces_cached_matches_atomically(tmp_path: P
with sqlite3.connect(state_dir / "workbench.sqlite3") as connection:
assert connection.execute("SELECT COUNT(*) FROM scan_comparisons").fetchone() == (1,)
assert connection.execute("SELECT COUNT(*) FROM scan_comparison_matches").fetchone() == (0,)
- shown = run_workbench(state_dir, "get-scan", "--scan-id", before["scanId"])["scan"]["findings"][
- 0
- ]
+ shown = get_scan(state_dir, before["scanId"])["scan"]["findings"][0]
assert "matches" not in shown
@@ -1378,18 +1284,8 @@ def test_cli_scan_comparison_requires_complete_matching_path_coverage(tmp_path:
other_scope = create_cli_scan(state_dir, root, repository, finding=False, paths=["tests"])
deep = create_cli_scan(state_dir, root, repository, mode="deep")
explicit_root = create_cli_scan(state_dir, root, repository, paths=["."])
- assert (
- run_workbench(state_dir, "get-scan", "--scan-id", deep["scanId"])["scan"]["progress"][
- "status"
- ]
- == "complete"
- )
- assert (
- run_workbench(state_dir, "get-scan", "--scan-id", explicit_root["scanId"])["scan"][
- "progress"
- ]["status"]
- == "complete"
- )
+ assert get_scan(state_dir, deep["scanId"])["scan"]["progress"]["status"] == "complete"
+ assert get_scan(state_dir, explicit_root["scanId"])["scan"]["progress"]["status"] == "complete"
for later in (partial, other_scope):
compared = run_workbench(
diff --git a/plugins/codex-security/tests/test_workbench_scan_usage.py b/plugins/codex-security/tests/test_workbench_scan_usage.py
index f7db562dc..ea9fdeedd 100644
--- a/plugins/codex-security/tests/test_workbench_scan_usage.py
+++ b/plugins/codex-security/tests/test_workbench_scan_usage.py
@@ -13,11 +13,15 @@ from typing import Any
import pytest
from workbench_test_support import (
+ begin_deep_scan,
create_saved_workspace,
+ fail_scan,
initialize_git_repository,
mark_deep_coordinator_succeeded,
run_workbench,
+ scan_command,
start_delivered_scan,
+ upsert_deep_worker,
write_completed_contract,
)
@@ -122,7 +126,6 @@ def _token_event(
output_tokens: int,
*,
cached_input_tokens: int = 0,
- reasoning_output_tokens: int = 0,
) -> dict[str, Any]:
return _event(
timestamp,
@@ -135,7 +138,7 @@ def _token_event(
"cached_input_tokens": cached_input_tokens,
"cache_write_input_tokens": 0,
"output_tokens": output_tokens,
- "reasoning_output_tokens": reasoning_output_tokens,
+ "reasoning_output_tokens": 0,
"total_tokens": input_tokens + output_tokens,
}
},
@@ -215,7 +218,6 @@ def _counts(
input_tokens: int,
cached_input_tokens: int,
output_tokens: int,
- reasoning_output_tokens: int = 0,
*,
cache_write_input_tokens: int = 0,
) -> dict[str, int]:
@@ -224,7 +226,7 @@ def _counts(
"cachedInputTokens": cached_input_tokens,
"cacheWriteInputTokens": cache_write_input_tokens,
"outputTokens": output_tokens,
- "reasoningOutputTokens": reasoning_output_tokens,
+ "reasoningOutputTokens": 0,
"totalTokens": input_tokens + output_tokens,
}
@@ -248,12 +250,8 @@ def _complete_scan(fixture: ScanFixture) -> dict[str, Any]:
fixture.target,
**options,
)
- return run_workbench(
- fixture.state_dir,
- "complete-scan",
- "--scan-id",
- fixture.scan_id,
- environment=fixture.environment,
+ return scan_command(
+ fixture.state_dir, "complete-scan", fixture.scan_id, environment=fixture.environment
)
@@ -538,10 +536,8 @@ def test_completion_counts_deep_sdk_workers_and_descendants(tmp_path: Path) -> N
"CODEX_SQLITE_HOME": str(tmp_path / "codex-sqlite"),
"CODEX_STATE_DB": "",
}
- deep = run_workbench(
+ deep = begin_deep_scan(
state_dir,
- "begin-deep-scan",
- "--thread-id",
"scan-parent",
"--target-path",
str(target),
@@ -572,20 +568,13 @@ def test_completion_counts_deep_sdk_workers_and_descendants(tmp_path: Path) -> N
artifact.mkdir(parents=True)
prompt = artifact / "prompt.md"
prompt.write_text("Review the fixture target.\n", encoding="utf-8")
- run_workbench(
+ upsert_deep_worker(
state_dir,
- "upsert-deep-scan-worker",
- "--scan-id",
scan_id,
- "--worker-id",
str(uuid.uuid4()),
- "--kind",
"discovery",
- "--status",
"running",
- "--prompt-path",
str(prompt),
- "--artifact-dir",
str(artifact),
"--sdk-thread-id",
"sdk-worker",
@@ -633,10 +622,9 @@ def test_completion_preserves_explicit_legacy_cost(tmp_path: Path) -> None:
write_completed_contract(
fixture.scan_dir, fixture.scan_id, fixture.target, relative_path="app.py"
)
- completed = run_workbench(
+ completed = scan_command(
fixture.state_dir,
"complete-scan",
- "--scan-id",
fixture.scan_id,
"--cost-json",
json.dumps(cost),
@@ -655,10 +643,9 @@ def test_usage_is_returned_by_completion_without_an_extra_command(tmp_path: Path
[],
)
assert _complete_scan(fixture)["scan"]["usage"]["totalTokens"] == 16
- extra_command = run_workbench(
+ extra_command = scan_command(
fixture.state_dir,
"get-scan-usage",
- "--scan-id",
fixture.scan_id,
check=False,
environment=fixture.environment,
@@ -675,14 +662,8 @@ def test_failed_scan_preserves_legacy_failure_behavior(tmp_path: Path) -> None:
{"scan-parent": _rollout(tmp_path, "scan-parent", [_token_event(counted, 21, 8)])},
[],
)
- failed = run_workbench(
- fixture.state_dir,
- "fail-scan",
- "--scan-id",
- fixture.scan_id,
- "--message",
- "Fixture failure.",
- environment=fixture.environment,
+ failed = fail_scan(
+ fixture.state_dir, fixture.scan_id, "Fixture failure.", environment=fixture.environment
)["scan"]
assert failed["progress"]["status"] == "failed"
assert "usage" not in failed