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

refactor(plugin): simplify workbench owner calls (#1205)

mldangelo-oai37cd0ed191f0
diff --git a/plugins/codex-security/mcp-app/src/python_command.ts b/plugins/codex-security/mcp-app/src/python_command.ts
index e8af5e646..76bba7fec 100644
--- a/plugins/codex-security/mcp-app/src/python_command.ts
+++ b/plugins/codex-security/mcp-app/src/python_command.ts
@@ -67,8 +67,7 @@ export async function isUsablePythonExecutable(
platform: PythonPlatform,
): Promise<boolean> {
try {
- const candidateStat = await fs.stat(candidate);
- if (!candidateStat.isFile()) {
+ if (!(await fs.stat(candidate)).isFile()) {
return false;
}
if (platform !== "win32") {
diff --git a/plugins/codex-security/scripts/config_preflight.py b/plugins/codex-security/scripts/config_preflight.py
index c30614a08..a0e7127e2 100644
--- a/plugins/codex-security/scripts/config_preflight.py
+++ b/plugins/codex-security/scripts/config_preflight.py
@@ -17,6 +17,10 @@ try:
except ModuleNotFoundError: # pragma: no cover - Python 3.10 only
import tomli as tomllib
+# Some plugin hosts launch Python with safe-path isolation enabled.
+sys.path.insert(0, str(Path(__file__).resolve().parent))
+from workbench_constants import positive_int
+
PLUGIN_ROOT = Path(__file__).resolve().parents[1]
DEFAULT_REGISTRY = PLUGIN_ROOT / "preflight" / "capability-profiles.toml"
DEFAULT_CODEX_HOME = Path(os.environ.get("CODEX_HOME", "~/.codex")).expanduser()
@@ -147,13 +151,6 @@ def parse_bool(value: str) -> bool:
raise ValueError(f"expected true or false, got {value!r}")
-def positive_int(value: str) -> int:
- parsed = int(value)
- if parsed < 1:
- raise argparse.ArgumentTypeError("expected a positive integer")
- return parsed
-
-
def parse_assignment(raw: str) -> tuple[str, str]:
key, separator, value = raw.partition("=")
if not separator or not key or not value:
@@ -865,7 +862,6 @@ def evaluate(args: argparse.Namespace) -> dict[str, Any]:
if path in project_config_paths_set:
# Codex strips project-local profile selection and definitions before
# resolving the merged config. Mirror that denylist here.
- config = dict(config)
config.pop("profile", None)
config.pop("profiles", None)
config_layers.append((path, config))
diff --git a/plugins/codex-security/scripts/deep_scan_workbench.py b/plugins/codex-security/scripts/deep_scan_workbench.py
index 1ccb1a8f7..bfe93f061 100644
--- a/plugins/codex-security/scripts/deep_scan_workbench.py
+++ b/plugins/codex-security/scripts/deep_scan_workbench.py
@@ -150,10 +150,6 @@ def dependencies() -> Any:
return _dependencies
-def now() -> str:
- return dependencies().now()
-
-
def _bounded_error_text(message: str, maximum: int) -> str:
if len(message) <= maximum:
return message
@@ -188,18 +184,6 @@ def _parse_timestamp(value: str) -> datetime:
return datetime.fromisoformat(value)
-def require_scan(connection: sqlite3.Connection, scan_id: str) -> sqlite3.Row:
- return dependencies().require_scan(connection, scan_id)
-
-
-def require_workspace(connection: sqlite3.Connection, workspace_id: str) -> sqlite3.Row:
- return dependencies().require_workspace(connection, workspace_id)
-
-
-def require_canonical_scan_directory(scan_dir: Path) -> Path:
- return dependencies().require_canonical_scan_directory(scan_dir)
-
-
def require_deep_scan_run(connection: sqlite3.Connection, scan_id: str) -> sqlite3.Row:
scan_id = require_uuid(scan_id, "scan-id")
row = connection.execute(
@@ -211,7 +195,7 @@ def require_deep_scan_run(connection: sqlite3.Connection, scan_id: str) -> sqlit
def deep_scan_deadline_reached(run: sqlite3.Row) -> bool:
- elapsed = _parse_timestamp(now()) - _parse_timestamp(str(run["created_at"]))
+ elapsed = _parse_timestamp(dependencies().now()) - _parse_timestamp(str(run["created_at"]))
return elapsed.total_seconds() / 3600 >= run["max_time_hours"]
@@ -234,8 +218,8 @@ def require_deep_scan_ready_for_parent_completion(
def require_owned_scan(
connection: sqlite3.Connection, scan_id: str, thread_id: str
) -> tuple[sqlite3.Row, sqlite3.Row]:
- scan = require_scan(connection, scan_id)
- workspace = require_workspace(connection, scan["workspace_id"])
+ scan = dependencies().require_scan(connection, scan_id)
+ workspace = dependencies().require_workspace(connection, scan["workspace_id"])
owner = optional_text(thread_id, maximum=512)
if owner is None:
raise SystemExit("thread-id is required.")
@@ -257,7 +241,7 @@ def deep_scan_path(
raise SystemExit(f"{label} must be an absolute path inside the scan directory.")
try:
resolved = supplied.resolve(strict=True)
- scan_dir = require_canonical_scan_directory(Path(scan["scan_dir"]))
+ scan_dir = dependencies().require_canonical_scan_directory(Path(scan["scan_dir"]))
resolved.relative_to(scan_dir)
except (OSError, RuntimeError, ValueError) as exc:
raise SystemExit(f"{label} must be an existing path inside the scan directory.") from exc
@@ -352,7 +336,7 @@ def canonical_discovery_artifacts(scan: sqlite3.Row) -> dict[str, str]:
def deep_scan_state(connection: sqlite3.Connection, scan_id: str) -> dict[str, Any]:
run = require_deep_scan_run(connection, scan_id)
- scan = require_scan(connection, run["scan_id"])
+ scan = dependencies().require_scan(connection, run["scan_id"])
worker_rows = connection.execute(
"""
SELECT *
@@ -627,8 +611,8 @@ def begin_deep_scan_for_scan(
args: argparse.Namespace,
) -> dict[str, Any]:
scan_id = require_uuid(scan_id, "scan-id")
- candidate = require_scan(connection, scan_id)
- workspace = require_workspace(connection, candidate["workspace_id"])
+ candidate = dependencies().require_scan(connection, scan_id)
+ workspace = dependencies().require_workspace(connection, candidate["workspace_id"])
if (
candidate["mode"] == "deep"
and candidate["status"] == "running"
@@ -642,7 +626,7 @@ def begin_deep_scan_for_scan(
args.claim_token,
error_message="Deep Scan orchestration is owned by another continuation.",
)
- timestamp = now()
+ timestamp = dependencies().now()
with connection:
claimed_workspace = connection.execute(
"UPDATE workspaces SET thread_id = ?, updated_at = ? "
@@ -694,7 +678,7 @@ def begin_deep_scan_for_scan(
args.claim_token,
error_message="Deep Scan orchestration is owned by another continuation.",
)
- ensure_deep_scan_run(connection, scan, config, workflow_version, now())
+ ensure_deep_scan_run(connection, scan, config, workflow_version, dependencies().now())
connection.commit()
except BaseException:
connection.rollback()
@@ -736,7 +720,9 @@ def begin_deep_scan_for_target(
workflow_version = optional_text(args.workflow_version, maximum=256)
if workflow_version is None:
raise SystemExit("workflow-version is required.")
- ensure_deep_scan_run(connection, existing, config, workflow_version, now())
+ ensure_deep_scan_run(
+ connection, existing, config, workflow_version, dependencies().now()
+ )
connection.commit()
return deep_scan_result(
connection,
@@ -787,7 +773,7 @@ def begin_deep_scan_for_target(
reasoning_effort = optional_text(args.reasoning_effort, maximum=32)
workspace_id = str(uuid.uuid4())
scan_id = str(uuid.uuid4())
- timestamp = now()
+ timestamp = dependencies().now()
target_id = dependencies().ensure_security_target(connection, target_path)
scan_dir = Path(
tempfile.mkdtemp(
@@ -857,7 +843,7 @@ def begin_deep_scan_for_target(
"UPDATE workspaces SET active_scan_id = ?, updated_at = ? WHERE id = ?",
(scan_id, timestamp, workspace_id),
)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
ensure_deep_scan_run(connection, scan, config, workflow_version, timestamp)
connection.commit()
except BaseException:
@@ -953,7 +939,7 @@ def claim_deep_scan_coordinator_locked(
error_message="Deep Scan orchestration is owned by another continuation.",
)
run, _ = require_running_deep_scan(connection, scan_id)
- timestamp = now()
+ timestamp = dependencies().now()
if args.coordinator_generation is not None:
require_current_coordinator(run, args)
disposition = "claimed"
@@ -1065,7 +1051,7 @@ def recover_expired_coordinator(
def recover_candidate_ledger_publication(connection: sqlite3.Connection, scan_id: str) -> None:
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
ledger = Path(scan["scan_dir"]) / "artifacts" / "02_discovery" / "candidate_ledger.jsonl"
backups = sorted(
ledger.parent.glob(f".{ledger.name}.*.backup"),
@@ -1116,7 +1102,7 @@ def require_running_deep_scan(
connection: sqlite3.Connection, scan_id: str
) -> tuple[sqlite3.Row, sqlite3.Row]:
run = require_deep_scan_run(connection, scan_id)
- scan = require_scan(connection, run["scan_id"])
+ scan = dependencies().require_scan(connection, run["scan_id"])
if run["status"] != "running" or run["cancel_requested"]:
raise SystemExit("Only a running Deep Scan can update orchestration state.")
if scan["status"] != "running" or scan["canceled_at"] is not None:
@@ -1145,7 +1131,7 @@ def upsert_deep_scan_worker(
try:
run = require_deep_scan_run(connection, scan_id)
require_current_coordinator(run, args)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
existing = connection.execute(
"SELECT * FROM deep_scan_workers WHERE id = ?", (worker_id,)
).fetchone()
@@ -1188,7 +1174,7 @@ def upsert_deep_scan_worker(
if args.result_manifest_path
else None
)
- timestamp = now()
+ timestamp = dependencies().now()
if existing is None:
if args.kind == "dedup":
raise SystemExit("Create dedup workers with claim-deep-scan-dedup.")
@@ -1469,7 +1455,7 @@ def claim_deep_scan_dedup(
if minimum_inputs == 2
else "A Deep Scan dedup requires at least one buffered discovery result."
)
- timestamp = now()
+ timestamp = dependencies().now()
connection.execute(
"""
INSERT INTO deep_scan_workers (
@@ -1533,7 +1519,7 @@ def commit_deep_scan_dedup_locked(
try:
run = require_deep_scan_run(connection, scan_id)
require_current_coordinator(run, args)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
worker = require_deep_scan_worker(connection, worker_id)
if worker["scan_id"] != scan_id or worker["kind"] != "dedup":
raise SystemExit("Dedup worker does not belong to this Deep Scan.")
@@ -1595,7 +1581,7 @@ def commit_deep_scan_dedup_locked(
str(publication_copy),
canonical_candidate_ledger_path,
)
- timestamp = now()
+ timestamp = dependencies().now()
connection.execute(
"""
UPDATE deep_scan_workers
@@ -1662,7 +1648,7 @@ def finish_deep_scan_locked(
try:
run = require_deep_scan_run(connection, scan_id)
require_current_coordinator(run, args)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
manifest_path = (
deep_scan_output_path(scan, args.manifest_path, "Deep Scan coordinator manifest path")
if args.staged_manifest_path
@@ -1727,7 +1713,7 @@ def finish_deep_scan_locked(
AND reducers.status IN ('failed', 'canceled')
)
""",
- (now(), scan_id, scan_id),
+ (dependencies().now(), scan_id, scan_id),
)
buffered_worker_ids = [
row["id"]
@@ -1759,7 +1745,7 @@ def finish_deep_scan_locked(
if run["manifest_path"] is None:
connection.execute(
"UPDATE deep_scan_runs SET manifest_path = ?, updated_at = ? WHERE scan_id = ?",
- (manifest_path, now(), scan_id),
+ (manifest_path, dependencies().now(), scan_id),
)
connection.commit()
return deep_scan_result(connection, scan_id)
@@ -1853,7 +1839,7 @@ def finish_deep_scan_locked(
if args.terminal_reason == "saturated":
# Mark any remaining workers canceled, including those whose own
# cancellation writes failed, so they cannot block completion.
- cancel_active_workers(connection, scan_id, now())
+ cancel_active_workers(connection, scan_id, dependencies().now())
active_worker = connection.execute(
"""
SELECT 1 FROM deep_scan_workers
@@ -1893,7 +1879,7 @@ def finish_deep_scan_locked(
kind="file",
)
promotion = promote_staged_file(staged_manifest_path, manifest_path)
- timestamp = now()
+ timestamp = dependencies().now()
connection.execute(
"""
UPDATE deep_scan_runs
@@ -1932,7 +1918,7 @@ def fail_deep_scan_locked(
try:
run = require_deep_scan_run(connection, scan_id)
require_current_coordinator(run, args)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
manifest_path = None
if args.manifest_path:
manifest_path = (
@@ -1979,7 +1965,7 @@ def fail_deep_scan_locked(
kind="file",
)
promotion = promote_staged_file(staged_manifest_path, manifest_path)
- timestamp = now()
+ timestamp = dependencies().now()
connection.execute(
"""
UPDATE deep_scan_runs
@@ -2028,7 +2014,7 @@ def record_deep_scan_publication_failure(
try:
run = require_deep_scan_run(connection, scan_id)
require_current_coordinator(run, args)
- scan = require_scan(connection, scan_id)
+ scan = dependencies().require_scan(connection, scan_id)
if (
run["status"] not in {"failed", "canceled", "interrupted"}
or scan["status"] != "failed"
@@ -2040,7 +2026,7 @@ def record_deep_scan_publication_failure(
connection.commit()
return deep_scan_result(connection, scan_id)
if run["publication_error_message"] != message:
- timestamp = now()
+ timestamp = dependencies().now()
connection.execute(
"""
UPDATE deep_scan_runs
@@ -2061,7 +2047,7 @@ def clear_deep_scan_publication_failure(connection: sqlite3.Connection, scan_id:
connection.execute(
"UPDATE deep_scan_runs SET publication_error_message = NULL, updated_at = ? "
"WHERE scan_id = ? AND publication_error_message IS NOT NULL",
- (now(), scan_id),
+ (dependencies().now(), scan_id),
)
diff --git a/plugins/codex-security/scripts/filesystem_identity.py b/plugins/codex-security/scripts/filesystem_identity.py
index dea3d3cb5..91d6c5346 100644
--- a/plugins/codex-security/scripts/filesystem_identity.py
+++ b/plugins/codex-security/scripts/filesystem_identity.py
@@ -22,9 +22,5 @@ def stored_filesystem_identity_matches(stored: object, current: int) -> bool:
return stored == serialize_filesystem_identity(current)
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/finalize_scan_contract.py b/plugins/codex-security/scripts/finalize_scan_contract.py
index 2c7181d0a..e89bd3d58 100644
--- a/plugins/codex-security/scripts/finalize_scan_contract.py
+++ b/plugins/codex-security/scripts/finalize_scan_contract.py
@@ -352,10 +352,6 @@ def write_threat_model_projection_if_possible(
return warning
-def _validate_report_output_paths(scan_dir: Path) -> None:
- _validate_scan_local_output_path(scan_dir, scan_dir / "report.md", "report.md")
-
-
def _json_bytes(payload: Any) -> bytes:
try:
encoded = json.dumps(payload, allow_nan=False, indent=2, sort_keys=True)
@@ -406,22 +402,6 @@ def _require_safe_json_string(value: str, context: str) -> None:
raise ContractError(f"{context}: expected well-formed Unicode JSON strings") from exc
-def _sha256_text(value: str) -> str:
- return hashlib.sha256(value.encode("utf-8")).hexdigest()
-
-
-def _sha256_bytes(value: bytes) -> str:
- return hashlib.sha256(value).hexdigest()
-
-
-def _sha256_file(path: Path) -> str:
- digest = hashlib.sha256()
- with path.open("rb") as handle:
- for chunk in iter(lambda: handle.read(1024 * 1024), b""):
- digest.update(chunk)
- return digest.hexdigest()
-
-
def _require_dict(payload: dict[str, Any], key: str, context: str) -> dict[str, Any]:
value = payload.get(key)
if not isinstance(value, dict):
@@ -508,10 +488,6 @@ def _validate_scan_local_output_path(scan_dir: Path, path: Path, relative_path:
raise ContractError(f"{relative_path}: expected a regular file")
-def _descriptor_relative_reads_available() -> bool:
- return os.open in os.supports_dir_fd and hasattr(os, "O_NOFOLLOW")
-
-
def _is_windows() -> bool:
return os.name == "nt"
@@ -610,7 +586,7 @@ def _open_scan_local_directory(root_fd: int, parts: tuple[str, ...], *, create:
def open_scan_local_file_descriptor(scan_dir: Path, relative_path: str, context: str) -> int:
scan_dir = _require_scan_directory(scan_dir)
relative_path = _require_portable_relative_path(relative_path, context)
- if not _descriptor_relative_reads_available():
+ if not (os.open in os.supports_dir_fd and hasattr(os, "O_NOFOLLOW")):
if not _is_windows():
raise ContractError("scan-local input requires descriptor-relative file operations")
try:
@@ -915,11 +891,7 @@ def _validate_target(target: dict[str, Any]) -> None:
_validate_remote(remote, "scan.target.remote")
if kind == "git_revision":
_require_str(target, "revision", "scan.target")
- elif kind == "git_worktree":
- _require_str(target, "snapshotDigest", "scan.target")
- elif kind == "git_diff":
- _require_str(target, "snapshotDigest", "scan.target")
- elif kind == "directory_snapshot":
+ elif kind in ("git_worktree", "git_diff", "directory_snapshot"):
_require_str(target, "snapshotDigest", "scan.target")
@@ -937,11 +909,11 @@ def _fingerprint(target_id: str, finding: dict[str, Any]) -> str:
if not SLUG_RE.fullmatch(rule_id):
raise ContractError("finding.ruleId: expected a stable lowercase rule slug")
material = "\0".join((FINGERPRINT_ALGORITHM, target_id, rule_id, anchor, instance))
- return f"{FINGERPRINT_ALGORITHM}:sha256:{_sha256_text(material)}"
+ return f"{FINGERPRINT_ALGORITHM}:sha256:{hashlib.sha256(material.encode('utf-8')).hexdigest()}"
def _stable_id(prefix: str, *parts: str) -> str:
- return f"{prefix}_{_sha256_text(chr(0).join(parts))[:24]}"
+ return f"{prefix}_{hashlib.sha256(chr(0).join(parts).encode('utf-8')).hexdigest()[:24]}"
def _validate_location(location: dict[str, Any], context: str) -> None:
@@ -2582,7 +2554,7 @@ def _artifact_record(
"mediaType": media_type,
"path": relative_path,
"sha256": (
- _sha256_bytes(contents)
+ hashlib.sha256(contents).hexdigest()
if contents is not None
else _sha256_scan_local_file(scan_dir, relative_path, relative_path)
),
@@ -2633,7 +2605,7 @@ def _validate_existing_seal(
expected_sha256 = _require_str(artifact, "sha256", context)
contents = (artifact_contents or {}).get(path)
actual_sha256 = (
- _sha256_bytes(contents)
+ hashlib.sha256(contents).hexdigest()
if contents is not None
else _sha256_scan_local_file(scan_dir, path, context)
)
@@ -3024,12 +2996,12 @@ def _prepare_scan_finalization(
_validate_manifest(manifest)
validate_against_schema(manifest, schema_dir / "scan-manifest.schema.json")
report_markdown_bytes = _generate_report_projection(manifest, findings, coverage)
- _validate_report_output_paths(scan_dir)
+ _validate_scan_local_output_path(scan_dir, scan_dir / "report.md", "report.md")
else:
findings_bytes = _contract_json_bytes("findings.json", findings)
coverage_bytes = _contract_json_bytes("coverage.json", coverage)
report_markdown_bytes = _generate_report_projection(manifest, findings, coverage)
- _validate_report_output_paths(scan_dir)
+ _validate_scan_local_output_path(scan_dir, scan_dir / "report.md", "report.md")
scan["artifacts"] = [
_artifact_record(scan_dir, "findings.json", "application/json", findings_bytes),
_artifact_record(scan_dir, "coverage.json", "application/json", coverage_bytes),
diff --git a/plugins/codex-security/scripts/generate_rank_input.py b/plugins/codex-security/scripts/generate_rank_input.py
index 02b1f5924..b589cb061 100644
--- a/plugins/codex-security/scripts/generate_rank_input.py
+++ b/plugins/codex-security/scripts/generate_rank_input.py
@@ -32,11 +32,12 @@ import re
import subprocess
import sys
from collections import Counter
-from collections.abc import Callable
+from collections.abc import Callable, KeysView
from pathlib import Path
# Some plugin hosts launch Python with safe-path isolation enabled.
sys.path.insert(0, str(Path(__file__).resolve().parent))
+from generate_in_scope_files import windows_stream_component
from rank_preview import (
DEFAULT_PREVIEW_BYTES,
TEXT_CODE_EXTENSIONS,
@@ -129,7 +130,6 @@ RANK_POOL_STRATEGY = "round_robin"
RANK_POOL_WORKER_CAP = 6
JsonRow = dict[str, object]
RowValidator = Callable[[JsonRow, Path, int], None]
-RankWorkerAssignment = tuple[int, list[str], list[str]]
def parse_args() -> argparse.Namespace:
@@ -299,17 +299,6 @@ def path_is_diff_excluded(path: Path) -> bool:
return path_is_excluded(path)
-def windows_stream_component(path: Path) -> str | None:
- """Return the first NTFS alternate-data-stream component."""
-
- if os.name != "nt":
- return None
- return next(
- (component for component in path.parts if component != path.anchor and ":" in component),
- None,
- )
-
-
def resolve_scope(
repo: Path,
scope: str,
@@ -394,12 +383,11 @@ def load_jsonl(path: Path, label: str, validator: RowValidator) -> list[JsonRow]
if not raw_line.strip():
raise SystemExit(f"{path}:{line_number}: blank JSONL rows are not allowed")
try:
- parsed: object = json.loads(raw_line)
+ row: object = json.loads(raw_line)
except json.JSONDecodeError as exc:
raise SystemExit(f"{path}:{line_number}: invalid JSON: {exc.msg}") from exc
- if not isinstance(parsed, dict):
+ if not isinstance(row, dict):
raise SystemExit(f"{path}:{line_number}: expected a JSON object")
- row = {str(key): value for key, value in parsed.items()}
validator(row, path, line_number)
rows.append(row)
return rows
@@ -448,16 +436,12 @@ def validate_rank_output_row(row: JsonRow, path: Path, line_number: int) -> None
require_string(row, "reason", path, line_number, allow_empty=False)
-def require_unique_paths(rows: list[JsonRow], label: str) -> None:
- seen: set[str] = set()
- duplicates: set[str] = set()
- for row in rows:
- path = str(row["path"])
- if path in seen:
- duplicates.add(path)
- seen.add(path)
- if duplicates:
- raise SystemExit(f"{label} contains duplicate paths: {sorted(duplicates)}")
+def require_unique_paths(rows: list[JsonRow], label: str) -> KeysView[str]:
+ counts = Counter(str(row["path"]) for row in rows)
+ if len(counts) < len(rows):
+ duplicates = sorted(path for path, count in counts.items() if count > 1)
+ raise SystemExit(f"{label} contains duplicate paths: {duplicates}")
+ return counts.keys()
def make_repo_rank_input(args: argparse.Namespace) -> None:
@@ -710,9 +694,8 @@ def make_diff_rank_input(args: argparse.Namespace) -> None:
for path, status in changed:
rel = path.relative_to(repo)
- if status == "D":
- preview = ""
- elif args.mode == "revisions":
+ preview = ""
+ if status != "D" and args.mode == "revisions":
content = revision_blobs[rel]
if content is None:
raise SystemExit(
@@ -721,9 +704,7 @@ def make_diff_rank_input(args: argparse.Namespace) -> None:
preview, is_binary = preview_for_bytes(rel, content, args.preview_bytes)
if is_binary:
continue
- elif path.is_symlink():
- preview = ""
- elif path.is_file():
+ elif status != "D" and not path.is_symlink() and path.is_file():
try:
path.resolve(strict=True).relative_to(repo)
except (OSError, ValueError):
@@ -732,8 +713,6 @@ def make_diff_rank_input(args: argparse.Namespace) -> None:
preview, is_binary = preview_for(path, args.preview_bytes)
if is_binary:
continue
- else:
- preview = ""
rows.append({"path": rel.as_posix(), "area": args.area, "preview": preview})
rows.sort(key=lambda row: str(row["path"]))
@@ -757,8 +736,7 @@ def make_rank_shards(args: argparse.Namespace) -> None:
raise SystemExit(f"Rank shard directory already contains shard files: {output_dir}")
shard_count = 0
- for start in range(0, len(rows), args.max_rows):
- shard_count += 1
+ for shard_count, start in enumerate(range(0, len(rows), args.max_rows), start=1):
shard_path = output_dir / f"rank-shard-{shard_count:04d}.input.jsonl"
write_jsonl(shard_path, rows[start : start + args.max_rows])
@@ -858,7 +836,7 @@ def load_rank_pool_plan(plan_path: Path) -> tuple[dict[str, object], bytes]:
raise SystemExit(f"{plan_path}: invalid JSON: {exc.msg}") from exc
if not isinstance(payload, dict):
raise SystemExit(f"{plan_path}: expected a JSON object")
- return {str(key): value for key, value in payload.items()}, plan_bytes
+ return payload, plan_bytes
def require_integer(value: object, label: str, *, minimum: int) -> int:
@@ -875,24 +853,13 @@ def require_string_list(value: object, label: str) -> list[str]:
return [item for item in value if isinstance(item, str)]
-def assignment_differences(
- assigned_names: list[str], expected_names: list[str]
-) -> tuple[list[str], list[str], list[str]]:
- counts = Counter(assigned_names)
- duplicates = sorted(name for name, count in counts.items() if count > 1)
- assigned = set(assigned_names)
- expected = set(expected_names)
- return sorted(expected - assigned), duplicates, sorted(assigned - expected)
-
-
def validate_rank_pool_plan(
plan_path: Path, shard_dir: Path
-) -> tuple[list[Path], list[str], list[RankWorkerAssignment], bytes]:
+) -> tuple[list[Path], list[tuple[list[str], list[str]]], bytes]:
require_plan_shard_dir(plan_path, shard_dir)
require_no_misplaced_rank_shards(plan_path)
input_shards = discover_input_shards(shard_dir)
input_names = [path.name for path in input_shards]
- output_names = [output_name_for(name) for name in input_names]
plan, plan_bytes = load_rank_pool_plan(plan_path)
expected_fields = {
"schema_version",
@@ -940,15 +907,13 @@ def validate_rank_pool_plan(
f"{plan_path}: workers must contain exactly {worker_count} worker assignments"
)
- assigned_inputs: list[str] = []
- assigned_outputs: list[str] = []
- parsed_workers: list[RankWorkerAssignment] = []
+ counts: Counter[str] = Counter()
+ parsed_workers: list[tuple[list[str], list[str]]] = []
worker_fields = {"slot", "input_shards", "output_shards"}
- for worker_index, raw_worker in enumerate(workers):
+ for worker_index, worker in enumerate(workers):
label = f"{plan_path}: workers[{worker_index}]"
- if not isinstance(raw_worker, dict):
+ if not isinstance(worker, dict):
raise SystemExit(f"{label} must be a JSON object")
- worker = {str(key): value for key, value in raw_worker.items()}
if set(worker) != worker_fields:
raise SystemExit(
f"{label} fields do not match schema; "
@@ -965,47 +930,40 @@ def validate_rank_pool_plan(
expected_worker_outputs = [output_name_for(name) for name in worker_inputs]
if worker_outputs != expected_worker_outputs:
raise SystemExit(f"{label}.output_shards do not match its input_shards")
- assigned_inputs.extend(worker_inputs)
- assigned_outputs.extend(worker_outputs)
- parsed_workers.append((slot, worker_inputs, worker_outputs))
+ counts.update(worker_inputs)
+ parsed_workers.append((worker_inputs, worker_outputs))
- missing, duplicates, unexpected = assignment_differences(assigned_inputs, input_names)
+ # Canonical input names map one-to-one to the per-worker outputs checked above.
+ duplicates = sorted(name for name, count in counts.items() if count > 1)
+ expected = set(input_names)
+ missing = sorted(expected - counts.keys())
+ unexpected = sorted(counts.keys() - expected)
if missing or duplicates or unexpected:
raise SystemExit(
f"{plan_path}: pool plan must assign each input shard exactly once; "
f"missing={missing}; duplicates={duplicates}; unexpected={unexpected}"
)
- missing, duplicates, unexpected = assignment_differences(assigned_outputs, output_names)
- if missing or duplicates or unexpected:
- raise SystemExit(
- f"{plan_path}: pool plan must assign each output shard exactly once; "
- f"missing={missing}; duplicates={duplicates}; unexpected={unexpected}"
- )
-
- for worker_index, (_, worker_inputs, worker_outputs) in enumerate(parsed_workers):
+ for worker_index, (worker_inputs, _) in enumerate(parsed_workers):
expected_inputs = input_names[worker_index::worker_count]
- expected_outputs = output_names[worker_index::worker_count]
- if worker_inputs != expected_inputs or worker_outputs != expected_outputs:
+ if worker_inputs != expected_inputs:
raise SystemExit(
f"{plan_path}: worker slot {worker_index + 1} does not match the deterministic "
f"{RANK_POOL_STRATEGY} assignment"
)
- return input_shards, output_names, parsed_workers, plan_bytes
+ return input_shards, parsed_workers, plan_bytes
def validate_rank_worker_command(args: argparse.Namespace) -> None:
plan_path = Path(args.plan).expanduser()
shard_dir = Path(args.shard_dir).expanduser()
- _, _, workers, plan_bytes = validate_rank_pool_plan(plan_path, shard_dir)
+ _, workers, plan_bytes = validate_rank_pool_plan(plan_path, shard_dir)
slot = require_integer(args.slot, "--slot", minimum=1)
worker_count = len(workers)
if slot > worker_count:
raise SystemExit(f"--slot must be at most {worker_count}")
- assigned_slot, input_names, output_names = workers[slot - 1]
- if assigned_slot != slot:
- raise SystemExit(f"{plan_path}: worker assignment for slot {slot} is inconsistent")
+ input_names, output_names = workers[slot - 1]
row_count = 0
outputs_digest = hashlib.sha256()
@@ -1036,10 +994,10 @@ def validate_rank_worker_command(args: argparse.Namespace) -> None:
def validate_rank_pool_command(args: argparse.Namespace) -> None:
plan_path = Path(args.plan).expanduser()
shard_dir = Path(args.shard_dir).expanduser()
- input_shards, expected_output_names, workers, _ = validate_rank_pool_plan(plan_path, shard_dir)
+ input_shards, workers, _ = validate_rank_pool_plan(plan_path, shard_dir)
actual_output_names = {path.name for path in shard_dir.glob(SHARD_OUTPUT_GLOB)}
- expected_outputs = set(expected_output_names)
+ expected_outputs = {output_name_for(path.name) for path in input_shards}
if actual_output_names != expected_outputs:
missing = sorted(expected_outputs - actual_output_names)
unexpected = sorted(actual_output_names - expected_outputs)
@@ -1063,12 +1021,10 @@ def validate_rank_shard(
input_shard: Path, output_shard: Path
) -> tuple[list[JsonRow], list[JsonRow]]:
shard_inputs = load_jsonl(input_shard, "Rank input shard", validate_rank_input_row)
- require_unique_paths(shard_inputs, f"Rank input shard {input_shard.name}")
+ expected_paths = require_unique_paths(shard_inputs, f"Rank input shard {input_shard.name}")
shard_outputs = load_jsonl(output_shard, "Rank output shard", validate_rank_output_row)
- require_unique_paths(shard_outputs, f"Rank output shard {output_shard.name}")
+ actual_paths = require_unique_paths(shard_outputs, f"Rank output shard {output_shard.name}")
- expected_paths = {str(row["path"]) for row in shard_inputs}
- actual_paths = {str(row["path"]) for row in shard_outputs}
if expected_paths != actual_paths:
missing = sorted(expected_paths - actual_paths)
unknown = sorted(actual_paths - expected_paths)
diff --git a/plugins/codex-security/scripts/rank_preview.py b/plugins/codex-security/scripts/rank_preview.py
index 76084a998..95e54e4b6 100644
--- a/plugins/codex-security/scripts/rank_preview.py
+++ b/plugins/codex-security/scripts/rank_preview.py
@@ -679,8 +679,7 @@ def strip_leading_annotations(original: str, masked: str) -> tuple[str, str, lis
def brace_language_outline(text: str, suffix: str) -> list[str]:
original_lines = text.splitlines()
masked_lines = mask_c_style_source(text, suffix).splitlines()
- outline: list[str] = []
- seen: set[str] = set()
+ outline: dict[str, None] = {}
type_stack: list[tuple[str, int]] = []
function_depths: list[int] = []
blocked_depths: list[int] = []
@@ -691,9 +690,7 @@ def brace_language_outline(text: str, suffix: str) -> list[str]:
def add(value: str) -> None:
decorated = f"{' '.join(pending_annotations[-2:])} {value}".strip()
- if decorated not in seen:
- seen.add(decorated)
- outline.append(decorated)
+ outline[decorated] = None
def next_lines_open_body(line_index: int) -> bool:
inspected = 0
@@ -805,20 +802,14 @@ def brace_language_outline(text: str, suffix: str) -> list[str]:
type_stack.pop()
while blocked_depths and depth < blocked_depths[-1]:
blocked_depths.pop()
- return outline
+ return list(outline)
def ruby_outline(text: str) -> list[str]:
- outline: list[str] = []
- seen: set[str] = set()
+ outline: dict[str, None] = {}
type_stack: list[tuple[int, str]] = []
function_indents: list[int] = []
- def add(value: str) -> None:
- if value not in seen:
- seen.add(value)
- outline.append(value)
-
for raw_line in text.splitlines():
stripped = raw_line.strip()
if not stripped or stripped.startswith("#"):
@@ -832,7 +823,7 @@ def ruby_outline(text: str) -> list[str]:
type_match = re.match(r"^(class|module)\s+([A-Z]\w*(?:::[A-Z]\w*)*)", stripped)
if type_match and not function_indents:
kind, name = type_match.groups()
- add(f"{kind} {name}")
+ outline[f"{kind} {name}"] = None
type_stack.append((indent, name))
continue
@@ -840,21 +831,15 @@ def ruby_outline(text: str) -> list[str]:
if function_match and not function_indents:
name = function_match.group(1)
if type_stack:
- add(f"method {type_stack[-1][1]}.{name}")
+ outline[f"method {type_stack[-1][1]}.{name}"] = None
else:
- add(f"function {name}")
+ outline[f"function {name}"] = None
function_indents.append(indent)
- return outline
+ return list(outline)
def simple_language_outline(text: str, suffix: str) -> list[str]:
- outline: list[str] = []
- seen: set[str] = set()
-
- def add(value: str) -> None:
- if value not in seen:
- seen.add(value)
- outline.append(value)
+ outline: dict[str, None] = {}
for raw_line in text.splitlines():
line = compact_preview_line(raw_line)
@@ -864,61 +849,63 @@ def simple_language_outline(text: str, suffix: str) -> list[str]:
if suffix == ".py":
match = re.match(r"^(async\s+)?def\s+([A-Za-z_]\w*)\s*\(", line)
if match:
- add(f"{'async function' if match.group(1) else 'function'} {match.group(2)}")
+ outline[
+ f"{'async function' if match.group(1) else 'function'} {match.group(2)}"
+ ] = None
continue
match = re.match(r"^class\s+([A-Za-z_]\w*)", line)
if match:
- add(f"class {match.group(1)}")
+ outline[f"class {match.group(1)}"] = None
elif suffix in {".ex", ".exs"}:
match = re.match(r"^defmodule\s+([A-Za-z_]\w*(?:\.[A-Za-z_]\w*)*)", line)
if match:
- add(f"module {match.group(1)}")
+ outline[f"module {match.group(1)}"] = None
continue
match = re.match(r"^(?:def|defp|defmacro)\s+([A-Za-z_]\w*[!?]?)", line)
if match:
- add(f"function {match.group(1)}")
+ outline[f"function {match.group(1)}"] = None
elif suffix == ".clj":
match = re.match(
r"^\((defn|defmacro|defprotocol|defrecord|deftype|defmulti|defmethod)\s+([^\s)]+)",
line,
)
if match:
- add(f"{match.group(1)} {match.group(2)}")
+ outline[f"{match.group(1)} {match.group(2)}"] = None
elif suffix == ".lua":
match = re.match(r"^(?:local\s+)?function\s+([A-Za-z_]\w*(?:[.:][A-Za-z_]\w*)*)", line)
if match:
- add(f"function {match.group(1)}")
+ outline[f"function {match.group(1)}"] = None
elif suffix == ".sh":
match = re.match(r"^(?:function\s+)?([A-Za-z_]\w*)\s*\(\s*\)\s*\{?", line)
if match:
- add(f"function {match.group(1)}")
+ outline[f"function {match.group(1)}"] = None
elif suffix == ".hs":
match = re.match(r"^(data|newtype|type|class)\s+([A-Z]\w*)", line)
if match:
- add(f"{match.group(1)} {match.group(2)}")
+ outline[f"{match.group(1)} {match.group(2)}"] = None
continue
match = re.match(r"^([a-z_]\w*)\s*::", line)
if match:
- add(f"function {match.group(1)}")
+ outline[f"function {match.group(1)}"] = None
elif suffix == ".proto":
match = re.match(r"^(message|service|enum)\s+([A-Za-z_]\w*)", line)
if match:
- add(f"{match.group(1)} {match.group(2)}")
+ outline[f"{match.group(1)} {match.group(2)}"] = None
continue
match = re.match(r"^rpc\s+([A-Za-z_]\w*)\s*\(", line)
if match:
- add(f"rpc {match.group(1)}")
+ outline[f"rpc {match.group(1)}"] = None
elif suffix == ".graphql":
match = re.match(
r"^(type|interface|input|enum|union|scalar|directive)\s+([A-Za-z_]\w*)",
line,
)
if match:
- add(f"{match.group(1)} {match.group(2)}")
+ outline[f"{match.group(1)} {match.group(2)}"] = None
continue
match = re.match(r"^(query|mutation|subscription)\s+([A-Za-z_]\w*)", line)
if match:
- add(f"{match.group(1)} {match.group(2)}")
+ outline[f"{match.group(1)} {match.group(2)}"] = None
elif suffix == ".sql":
match = re.match(
r"^CREATE\s+(?:OR\s+REPLACE\s+)?(FUNCTION|PROCEDURE|TABLE|VIEW|TRIGGER)\s+"
@@ -927,16 +914,16 @@ def simple_language_outline(text: str, suffix: str) -> list[str]:
flags=re.IGNORECASE,
)
if match:
- add(f"{match.group(1).lower()} {match.group(2)}")
+ outline[f"{match.group(1).lower()} {match.group(2)}"] = None
elif suffix in {".yaml", ".yml"}:
match = re.match(r"^([A-Za-z0-9_.-]+):(?:\s|$)", raw_line)
if match:
- add(f"key {match.group(1)}")
+ outline[f"key {match.group(1)}"] = None
elif suffix in {".toml", ".cfg"}:
match = re.match(r"^\[([^]]+)]$", line)
if match:
- add(f"section {match.group(1)}")
- return outline
+ outline[f"section {match.group(1)}"] = None
+ return list(outline)
def json_outline(text: str) -> list[str]:
@@ -996,12 +983,8 @@ def preview_for_bytes(path: Path, data: bytes, preview_bytes: int) -> tuple[str,
return fit_preview_lines(preview_lines, preview_bytes), False
-def main() -> None:
+if __name__ == "__main__":
parser = argparse.ArgumentParser(
description="Build structural or sampled previews for Codex Security rank inputs."
)
parser.parse_args()
-
-
-if __name__ == "__main__":
- main()
diff --git a/plugins/codex-security/scripts/report_projection.py b/plugins/codex-security/scripts/report_projection.py
index de1f1e093..d230ac173 100644
--- a/plugins/codex-security/scripts/report_projection.py
+++ b/plugins/codex-security/scripts/report_projection.py
@@ -63,10 +63,6 @@ def _cell(value: Any) -> str:
return _text(value, "none").replace("|", "\\|").replace("\n", "<br>")
-def _link_label(value: Any, fallback: str) -> str:
- return _cell(value) or _cell(fallback)
-
-
def _deep_report_id(finding: dict[str, Any]) -> str:
extensions = finding.get("extensions")
if isinstance(extensions, dict):
@@ -141,10 +137,6 @@ def _deep_title_parts(finding: dict[str, Any]) -> tuple[str, str | None]:
return normalized, None
-def _deep_finding_title(finding: dict[str, Any]) -> str:
- return _deep_title_parts(finding)[0]
-
-
def _deep_finding_groups(
findings: list[dict[str, Any]], writeup_paths: list[str | None]
) -> list[list[tuple[int, dict[str, Any], str | None]]]:
@@ -157,7 +149,7 @@ def _deep_finding_groups(
def _deep_group_titles(group: list[tuple[int, dict[str, Any], str | None]]) -> str:
titles: list[str] = []
for _, finding, _ in group:
- title = _cell(_deep_finding_title(finding))
+ title = _cell(_deep_title_parts(finding)[0])
if title not in titles:
titles.append(title)
return "<br>".join(titles)
@@ -197,7 +189,7 @@ def _deep_group_report_labels(
def _deep_group_report_links(group: list[tuple[int, dict[str, Any], str | None]]) -> str:
labels = _deep_group_report_labels(group)
return "<br>".join(
- f"[{_link_label(label, 'Unidentified report')}](#finding-{number})"
+ f"[{_cell(label)}](#finding-{number})"
for label, (number, _, _) in zip(labels, group, strict=True)
)
@@ -206,7 +198,7 @@ def _deep_group_writeup_links(group: list[tuple[int, dict[str, Any], str | None]
labels = _deep_group_report_labels(group)
links: list[str] = []
for label, (_, _, report_path) in zip(labels, group, strict=True):
- report_id = _link_label(label, "Unidentified report")
+ report_id = _cell(label)
links.append(
f"[Open {report_id}]({report_path})" if report_path else f"{report_id}: inline below"
)
@@ -282,15 +274,11 @@ def _section_code_evidence(
and isinstance(item.get("code"), str)
and item["code"].strip()
)
- unique: list[dict[str, Any]] = []
- seen: set[tuple[str, str]] = set()
+ unique: dict[tuple[str, str], dict[str, Any]] = {}
for item in resolved:
key = (str(item.get("id", "")), item["code"])
- if key in seen:
- continue
- seen.add(key)
- unique.append(item)
- return unique
+ unique.setdefault(key, item)
+ return list(unique.values())
def merged_root_cause(value: dict[str, Any]) -> tuple[str | None, Any]:
@@ -919,7 +907,7 @@ def build_report_markdown(
for number, (finding, report_path) in enumerate(
zip(findings, writeup_paths, strict=True), 1
):
- title = _link_label(finding["title"], "Untitled finding")
+ title = _cell(finding["title"])
finding_link = f"[{title}](#finding-{number})"
writeup_link = f"[Open report]({report_path})" if report_path else "inline below"
lines.append(
@@ -1084,11 +1072,7 @@ def generate_report_markdown(
return build_report_markdown(manifest, findings, coverage).encode("utf-8")
-def main() -> int:
+if __name__ == "__main__":
parser = argparse.ArgumentParser(description=__doc__)
parser.parse_args()
- return 0
-
-
-if __name__ == "__main__":
- raise SystemExit(main())
+ raise SystemExit(0)
diff --git a/plugins/codex-security/scripts/windows_scan_local_files.py b/plugins/codex-security/scripts/windows_scan_local_files.py
index c67c86eb6..3375178ef 100644
--- a/plugins/codex-security/scripts/windows_scan_local_files.py
+++ b/plugins/codex-security/scripts/windows_scan_local_files.py
@@ -686,9 +686,5 @@ def unlink_if_exists(scan_dir: Path, relative_path: str) -> None:
_mark_handle_for_deletion(handle.value)
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_cli.py b/plugins/codex-security/scripts/workbench_cli.py
index 18d21d9b4..ab38dd9de 100644
--- a/plugins/codex-security/scripts/workbench_cli.py
+++ b/plugins/codex-security/scripts/workbench_cli.py
@@ -22,6 +22,7 @@ from workbench_constants import (
PHASE_PROGRESS_UNITS,
PHASES,
REMEDIATION_UPDATE_STATES,
+ positive_int,
)
@@ -400,12 +401,5 @@ def parse_args(description: str) -> argparse.Namespace:
return parser.parse_args(arguments)
-def positive_int(value: str) -> int:
- parsed = int(value)
- if parsed < 1:
- raise argparse.ArgumentTypeError("expected a positive integer")
- return parsed
-
-
if __name__ == "__main__":
parse_args(__doc__)
diff --git a/plugins/codex-security/scripts/workbench_constants.py b/plugins/codex-security/scripts/workbench_constants.py
index 2019115f8..fc0a19c3c 100644
--- a/plugins/codex-security/scripts/workbench_constants.py
+++ b/plugins/codex-security/scripts/workbench_constants.py
@@ -70,9 +70,12 @@ GIT_REPOSITORY_ENVIRONMENT = (
EMPTY_GIT_TREE = "4b825dc642cb6eb9a060e54bf8d69288fbee4904"
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
+def positive_int(value: str) -> int:
+ parsed = int(value)
+ if parsed < 1:
+ raise argparse.ArgumentTypeError("expected a positive integer")
+ return parsed
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_db.py b/plugins/codex-security/scripts/workbench_db.py
index 05b6418c3..c1d08a3c5 100644
--- a/plugins/codex-security/scripts/workbench_db.py
+++ b/plugins/codex-security/scripts/workbench_db.py
@@ -126,7 +126,7 @@ from workbench_target import (
git_command,
git_output,
git_revision,
- git_submodule_paths,
+ git_submodule_entries,
git_target_metadata,
git_worktree_context,
remediation_checkout_snapshot,
@@ -399,18 +399,6 @@ def inspect_setup_values(
}
-def inspect_setup(args: argparse.Namespace) -> dict[str, Any]:
- return inspect_setup_values(
- args.target_path,
- args.scope,
- args.mode,
- args.diff_target_kind,
- args.diff_base_revision,
- args.diff_head_revision,
- args.diff_content_digest,
- )
-
-
def require_review_changes_target(target: Path) -> str:
revision = require_git_worktree_head(target)
repository_root = git_output(target, "rev-parse", "--show-toplevel")
@@ -922,18 +910,6 @@ def start_scan(connection: sqlite3.Connection, args: argparse.Namespace) -> dict
return workspace_state(connection, workspace["id"])
-def start_prompt_only_scan(
- connection: sqlite3.Connection, args: argparse.Namespace
-) -> dict[str, Any]:
- return _start_prompt_driven_scan(connection, args, headless_standard=False)
-
-
-def start_headless_standard_scan(
- connection: sqlite3.Connection, args: argparse.Namespace
-) -> dict[str, Any]:
- return _start_prompt_driven_scan(connection, args, headless_standard=True)
-
-
def _start_prompt_driven_scan(
connection: sqlite3.Connection, args: argparse.Namespace, *, headless_standard: bool
) -> dict[str, Any]:
@@ -1878,10 +1854,6 @@ def recover_scan_results(
return saved_results.recover_scan_results(_WORKBENCH_DB_CONTEXT, connection, args)
-def write_scan_draft(connection: sqlite3.Connection, args: argparse.Namespace) -> dict[str, Any]:
- return saved_results.write_scan_draft(_WORKBENCH_DB_CONTEXT, connection, args)
-
-
def fail_scan(connection: sqlite3.Connection, args: argparse.Namespace) -> dict[str, Any]:
return saved_results.fail_scan(_WORKBENCH_DB_CONTEXT, connection, args)
@@ -2450,7 +2422,7 @@ def require_reviewed_patch_applied(
git_dir = git_output(target, "rev-parse", "--absolute-git-dir")
if git_dir is None:
raise SystemExit("Could not inspect the selected Git working tree.")
- excluded += git_submodule_paths(target)
+ excluded += tuple(path for path, _ in git_submodule_entries(target))
with tempfile.TemporaryDirectory(prefix="codex-security-remediation-") as temporary:
reviewed_patch = Path(temporary) / "reviewed.patch"
digest = hashlib.sha256()
@@ -3079,13 +3051,10 @@ def finding_artifact_paths(scan_dir: Path, details: dict[str, Any]) -> list[str]
except OSError:
return artifacts
- directories_seen = 0
- for current_directory, directory_names, file_names in os.walk(
- poc_root, topdown=True, followlinks=False
+ for directories_seen, (current_directory, directory_names, file_names) in enumerate(
+ os.walk(poc_root, topdown=True, followlinks=False), start=1
):
- directories_seen += 1
if directories_seen > FINDING_ARTIFACT_DIRECTORIES_LIMIT:
- directory_names[:] = []
break
current_path = Path(current_directory)
directory_names[:] = [
@@ -3366,7 +3335,15 @@ def main() -> None:
print(json.dumps(result, allow_nan=False, sort_keys=True))
return
if args.command == "inspect-setup":
- result = inspect_setup(args)
+ result = inspect_setup_values(
+ args.target_path,
+ args.scope,
+ args.mode,
+ args.diff_target_kind,
+ args.diff_base_revision,
+ args.diff_head_revision,
+ args.diff_content_digest,
+ )
print(json.dumps(result, allow_nan=False, sort_keys=True))
return
if args.command in {"save-artifact", "read-artifact"}:
@@ -3395,9 +3372,9 @@ def main() -> None:
elif args.command == "start-scan":
result = start_scan(connection, args)
elif args.command == "start-prompt-only-scan":
- result = start_prompt_only_scan(connection, args)
+ result = _start_prompt_driven_scan(connection, args, headless_standard=False)
elif args.command == "start-headless-standard-scan":
- result = start_headless_standard_scan(connection, args)
+ result = _start_prompt_driven_scan(connection, args, headless_standard=True)
elif args.command == "begin-deep-scan":
result = deep_scan.begin_deep_scan(connection, args)
elif args.command == "get-deep-scan":
@@ -3498,7 +3475,7 @@ def main() -> None:
elif args.command == "recover-scan-results":
result = recover_scan_results(connection, args)
elif args.command == "write-scan-draft":
- result = write_scan_draft(connection, args)
+ result = saved_results.write_scan_draft(_WORKBENCH_DB_CONTEXT, connection, args)
elif args.command == "save-scan-artifact":
result = saved_results.save_scan_artifact(_WORKBENCH_DB_CONTEXT, connection, args)
elif args.command == "mark-handoff-delivered":
diff --git a/plugins/codex-security/scripts/workbench_native_indexes.py b/plugins/codex-security/scripts/workbench_native_indexes.py
index d3d458b3d..71291b280 100644
--- a/plugins/codex-security/scripts/workbench_native_indexes.py
+++ b/plugins/codex-security/scripts/workbench_native_indexes.py
@@ -192,10 +192,7 @@ def list_repositories(
) -> dict[str, Any]:
scans = scan_history.list_scans(connection)["scans"]
scans_by_id = {scan["scanId"]: scan for scan in scans}
- scan_count_by_target: dict[str, int] = {}
- for scan in scans:
- target_id = scan["targetId"]
- scan_count_by_target[target_id] = scan_count_by_target.get(target_id, 0) + 1
+ scan_count_by_target = dict(Counter(scan["targetId"] for scan in scans))
latest_scan_by_target: dict[str, dict[str, Any]] = {}
for row in connection.execute(
diff --git a/plugins/codex-security/scripts/workbench_progress.py b/plugins/codex-security/scripts/workbench_progress.py
index fb1b5303c..7793abf4d 100644
--- a/plugins/codex-security/scripts/workbench_progress.py
+++ b/plugins/codex-security/scripts/workbench_progress.py
@@ -16,15 +16,12 @@ from workbench_validation import optional_text, require_uuid, user_context_argum
MAX_PREFLIGHT_ISSUES = 32
-def _javascript_string_length(value: str) -> int:
- return len(value.encode("utf-16-le", errors="surrogatepass")) // 2
-
-
def _preflight_issue_text(value: Any, maximum: int, label: str) -> str:
if not isinstance(value, str):
raise SystemExit(f"Preflight issue {label} must be text.")
normalized = value.strip()
- if not normalized or _javascript_string_length(normalized) > maximum:
+ # Match JavaScript's UTF-16 code-unit length.
+ if not normalized or len(normalized.encode("utf-16-le", errors="surrogatepass")) // 2 > maximum:
raise SystemExit(f"Preflight issue {label} must contain 1 to {maximum} characters.")
return normalized
diff --git a/plugins/codex-security/scripts/workbench_remediation.py b/plugins/codex-security/scripts/workbench_remediation.py
index 275eb683a..2f5870d5c 100644
--- a/plugins/codex-security/scripts/workbench_remediation.py
+++ b/plugins/codex-security/scripts/workbench_remediation.py
@@ -193,9 +193,5 @@ def _cancel_generation(
)
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_scan_start.py b/plugins/codex-security/scripts/workbench_scan_start.py
index 7cb1f6217..b31f2543e 100644
--- a/plugins/codex-security/scripts/workbench_scan_start.py
+++ b/plugins/codex-security/scripts/workbench_scan_start.py
@@ -22,7 +22,7 @@ from workbench_target import (
git_revision,
worktree_content_digest,
)
-from workbench_validation import optional_text, user_text
+from workbench_validation import optional_text
def safe_segment(value: str) -> str:
@@ -168,7 +168,7 @@ def insert_running_scan(
) -> str:
revision = target_identity[0]
native_scan = scan_dir is None
- user_context = user_text(workspace["user_context"])
+ user_context = optional_text(workspace["user_context"])
if scan_dir is None:
scan_dir = Path(
tempfile.mkdtemp(
@@ -236,9 +236,5 @@ def insert_running_scan(
return scan_id
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_scan_usage.py b/plugins/codex-security/scripts/workbench_scan_usage.py
index 648924d92..d3e437e81 100644
--- a/plugins/codex-security/scripts/workbench_scan_usage.py
+++ b/plugins/codex-security/scripts/workbench_scan_usage.py
@@ -117,7 +117,6 @@ def collect_scan_usage(
return _unavailable_usage("scan_thread_unavailable", warnings=warnings)
total = _empty_token_usage()
- observed_thread_count = 0
accepted_thread_ids: set[str] = set()
excluded_thread_ids: set[str] = set()
for session in sessions:
@@ -151,17 +150,16 @@ def collect_scan_usage(
missing_thread_ids.add(session.thread_id)
continue
accepted_thread_ids.add(session.thread_id)
- observed_thread_count += 1
_add_token_usage(total, session_usage)
- if not observed_thread_count:
+ if not accepted_thread_ids:
return _unavailable_usage("scan_thread_unavailable", warnings=warnings)
result: dict[str, Any] = {
"coverage": "partial" if missing_thread_ids or warnings else "complete",
"source": "codex_rollout",
**total,
- "threadCount": observed_thread_count,
+ "threadCount": len(accepted_thread_ids),
}
if missing_thread_ids:
result["missingThreadCount"] = len(missing_thread_ids)
@@ -202,13 +200,11 @@ def _scan_root_thread_ids(
(scan["id"],),
)
)
- roots: list[str] = []
- seen: set[str] = set()
+ roots: dict[str, None] = {}
for candidate in candidates:
- if isinstance(candidate, str) and candidate.strip() and candidate not in seen:
- roots.append(candidate)
- seen.add(candidate)
- return roots
+ if isinstance(candidate, str) and candidate.strip():
+ roots[candidate] = None
+ return list(roots)
def _scan_execution_thread_ids(connection: sqlite3.Connection, scan: sqlite3.Row) -> list[str]:
diff --git a/plugins/codex-security/scripts/workbench_schema.py b/plugins/codex-security/scripts/workbench_schema.py
index 66baa2b4f..9efb0be4b 100644
--- a/plugins/codex-security/scripts/workbench_schema.py
+++ b/plugins/codex-security/scripts/workbench_schema.py
@@ -1543,9 +1543,5 @@ def sql_statements(script: str) -> list[str]:
return statements
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_source_excerpt.py b/plugins/codex-security/scripts/workbench_source_excerpt.py
index 06c4075c7..63e9fbe25 100644
--- a/plugins/codex-security/scripts/workbench_source_excerpt.py
+++ b/plugins/codex-security/scripts/workbench_source_excerpt.py
@@ -90,9 +90,5 @@ def safe_source_path(target: Path, relative_path: str) -> Path | None:
return path
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_target.py b/plugins/codex-security/scripts/workbench_target.py
index 25f8d8d30..2353cf019 100644
--- a/plugins/codex-security/scripts/workbench_target.py
+++ b/plugins/codex-security/scripts/workbench_target.py
@@ -454,10 +454,6 @@ def git_submodule_entries(target: Path) -> tuple[tuple[Path, str], ...]:
return tuple(entries)
-def git_submodule_paths(target: Path) -> tuple[Path, ...]:
- return tuple(path for path, _ in git_submodule_entries(target))
-
-
def require_clean_submodule_worktrees(target: Path) -> None:
for submodule, expected_revision in git_submodule_entries(target):
relative_path = str(submodule.relative_to(target))
@@ -872,9 +868,5 @@ def scan_target_warning(scan: sqlite3.Row) -> str | None:
return None
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_target_state.py b/plugins/codex-security/scripts/workbench_target_state.py
index 47d2ea6d3..7d24a5f20 100644
--- a/plugins/codex-security/scripts/workbench_target_state.py
+++ b/plugins/codex-security/scripts/workbench_target_state.py
@@ -55,9 +55,5 @@ def ensure_security_target(connection: sqlite3.Connection, target_path: str) ->
return target_id
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/scripts/workbench_validation.py b/plugins/codex-security/scripts/workbench_validation.py
index a65aff1e7..4459ed626 100644
--- a/plugins/codex-security/scripts/workbench_validation.py
+++ b/plugins/codex-security/scripts/workbench_validation.py
@@ -14,10 +14,10 @@ from typing import Any
# Some plugin hosts launch Python with safe-path isolation enabled.
sys.path.insert(0, str(Path(__file__).resolve().parent))
+import finalize_scan_contract as finalizer
+from workbench_scan_usage import _reject_nonstandard_json_number as reject_nonstandard_json_number
-
-def reject_non_finite_json(value: str) -> None:
- raise ValueError(f"non-finite JSON number {value!r} is not supported")
+reject_non_finite_json = finalizer._reject_non_finite_json
def require_uuid(value: str, label: str) -> str:
@@ -57,17 +57,9 @@ def require_close_note(close_reason: str | None, note: str | None) -> None:
raise SystemExit("Explain why this finding will not be fixed.")
-def user_text(value: str | None) -> str | None:
- return optional_text(value)
-
-
def user_context_argument(args: argparse.Namespace) -> str | None:
value = sys.stdin.read() if getattr(args, "user_context_stdin", False) else args.user_context
- return user_text(value)
-
-
-def reject_nonstandard_json_number(value: str) -> None:
- raise ValueError(f"invalid JSON number {value}")
+ return optional_text(value)
SCAN_USAGE_TOKEN_KEYS = (
@@ -195,9 +187,5 @@ def require_occurrence(connection: sqlite3.Connection, occurrence_id: str) -> sq
return row
-def main() -> None:
- argparse.ArgumentParser(description=__doc__).parse_args()
-
-
if __name__ == "__main__":
- main()
+ argparse.ArgumentParser(description=__doc__).parse_args()
diff --git a/plugins/codex-security/tests/test_capability_profiles.py b/plugins/codex-security/tests/test_capability_profiles.py
index 8a93ac3d8..a609d1efe 100644
--- a/plugins/codex-security/tests/test_capability_profiles.py
+++ b/plugins/codex-security/tests/test_capability_profiles.py
@@ -10,6 +10,8 @@ import sys
from pathlib import Path, PureWindowsPath
from types import SimpleNamespace
+import pytest
+
try:
import tomllib
except ModuleNotFoundError: # pragma: no cover - Python 3.10 only
@@ -52,23 +54,14 @@ def test_preflight_matches_windows_project_path_aliases(monkeypatch) -> None:
assert trust_level(layers, root) == "untrusted"
-def test_preflight_falls_back_to_tomli_without_stdlib_tomllib(monkeypatch) -> None:
- real_import = builtins.__import__
- monkeypatch.setitem(sys.modules, "tomli", tomllib)
-
- def import_without_tomllib(name, globals=None, locals=None, fromlist=(), level=0):
- if name == "tomllib":
- raise ModuleNotFoundError("No module named 'tomllib'", name="tomllib")
- return real_import(name, globals, locals, fromlist, level)
-
- monkeypatch.setattr(builtins, "__import__", import_without_tomllib)
-
- namespace = runpy.run_path(str(PREFLIGHT_SCRIPT), run_name="config_preflight_test")
-
- assert namespace["tomllib"] is tomllib
-
-
-def test_deep_scan_config_falls_back_to_tomli_without_stdlib_tomllib(monkeypatch) -> None:
+ ("script", "run_name"),
+ [
+ (PREFLIGHT_SCRIPT, "config_preflight_test"),
+ (DEEP_SCAN_CONFIG_SCRIPT, "deep_scan_config_test"),
+ ],
+)
+def test_config_falls_back_to_tomli_without_stdlib_tomllib(monkeypatch, script, run_name) -> None:
real_import = builtins.__import__
monkeypatch.setitem(sys.modules, "tomli", tomllib)
@@ -79,7 +72,7 @@ def test_deep_scan_config_falls_back_to_tomli_without_stdlib_tomllib(monkeypatch
monkeypatch.setattr(builtins, "__import__", import_without_tomllib)
- namespace = runpy.run_path(str(DEEP_SCAN_CONFIG_SCRIPT), run_name="deep_scan_config_test")
+ namespace = runpy.run_path(str(script), run_name=run_name)
assert namespace["tomllib"] is tomllib
diff --git a/plugins/codex-security/tests/test_finalize_scan_contract.py b/plugins/codex-security/tests/test_finalize_scan_contract.py
index 5864a7dbc..bdd7d1220 100644
--- a/plugins/codex-security/tests/test_finalize_scan_contract.py
+++ b/plugins/codex-security/tests/test_finalize_scan_contract.py
@@ -3,7 +3,6 @@ from __future__ import annotations
import copy
import csv
import hashlib
-import importlib.util
import io
import json
import os
@@ -14,25 +13,15 @@ import threading
import unittest
from datetime import datetime, timezone
from pathlib import Path
-from types import ModuleType
from unittest import mock
+from workbench_test_support import ScanFixtureTestCase, load_script, windows_file_backend
-def load_finalizer() -> ModuleType:
- script = Path(__file__).resolve().parent.parent / "scripts" / "finalize_scan_contract.py"
- spec = importlib.util.spec_from_file_location("finalize_scan_contract", script)
- if spec is None or spec.loader is None:
- raise RuntimeError(f"could not load {script}")
- module = importlib.util.module_from_spec(spec)
- spec.loader.exec_module(module)
- return module
-
-
-FINALIZER = load_finalizer()
+FINALIZER = load_script("finalize_scan_contract")
EXAMPLE_DIR = Path(__file__).resolve().parent.parent / "examples" / "completed-scan"
-class FinalizeScanContractTest(unittest.TestCase):
+class FinalizeScanContractTest(ScanFixtureTestCase):
def setUp(self) -> None:
self.temp_dir = tempfile.TemporaryDirectory()
self.scan_dir = Path(self.temp_dir.name).resolve()
@@ -127,9 +116,6 @@ class FinalizeScanContractTest(unittest.TestCase):
"deferred": [],
}
- def tearDown(self) -> None:
- self.temp_dir.cleanup()
-
def write_scan(self) -> None:
self.write_json("scan-manifest.json", self.manifest)
self.write_json("findings.json", self.findings)
@@ -166,9 +152,6 @@ The extraction root is not enforced.
path.parent.mkdir(parents=True, exist_ok=True)
path.write_text(json.dumps(payload, indent=2) + "\n", encoding="utf-8")
- def read_json(self, relative_path: str) -> dict[str, object]:
- return json.loads((self.scan_dir / relative_path).read_text(encoding="utf-8"))
-
def completion_binding(self) -> dict[str, object]:
return {
"scanId": "scan_001",
@@ -195,9 +178,7 @@ The extraction root is not enforced.
artifacts = manifest["scan"]["artifacts"]
for artifact in artifacts:
if artifact["path"] == relative_path:
- artifact["sha256"] = FINALIZER._sha256_bytes(
- (self.scan_dir / relative_path).read_bytes()
- )
+ artifact["sha256"] = self.sha256_file(relative_path)
break
else:
raise AssertionError(f"missing sealed artifact: {relative_path}")
@@ -375,7 +356,7 @@ The extraction root is not enforced.
manifest = self.read_json("scan-manifest.json")
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "findings.json":
- artifact["sha256"] = FINALIZER._sha256_file(self.scan_dir / "findings.json")
+ artifact["sha256"] = self.sha256_file("findings.json")
break
self.write_json("scan-manifest.json", manifest)
before = {
@@ -401,7 +382,7 @@ The extraction root is not enforced.
manifest = self.read_json("scan-manifest.json")
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "findings.json":
- artifact["sha256"] = FINALIZER._sha256_bytes(compact)
+ artifact["sha256"] = hashlib.sha256(compact).hexdigest()
break
self.write_json("scan-manifest.json", manifest)
self.assertGreater(len(FINALIZER._json_bytes(findings)), len(compact))
@@ -645,9 +626,7 @@ The extraction root is not enforced.
for artifact in manifest["scan"]["artifacts"]
if artifact["path"] == "coverage.json"
)
- coverage_artifact["sha256"] = FINALIZER._sha256_bytes(
- (self.scan_dir / "coverage.json").read_bytes()
- )
+ coverage_artifact["sha256"] = self.sha256_file("coverage.json")
self.write_json("scan-manifest.json", manifest)
with self.assertRaisesRegex(
@@ -1494,7 +1473,7 @@ The extraction root is not enforced.
manifest = self.read_json("scan-manifest.json")
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "findings.json":
- artifact["sha256"] = FINALIZER._sha256_file(self.scan_dir / "findings.json")
+ artifact["sha256"] = self.sha256_file("findings.json")
self.write_json("scan-manifest.json", manifest)
with self.assertRaisesRegex(FINALIZER.ContractError, "safe repository-relative"):
@@ -1662,28 +1641,7 @@ The extraction root is not enforced.
self.coverage["surfaces"][0]["disposition"] = "no_issue_found"
self.write_scan()
- backend = mock.Mock()
-
- def open_read_fd(scan_dir: Path, relative_path: str, _context: str) -> int:
- return os.open(scan_dir / relative_path, os.O_RDONLY)
-
- def atomic_write(
- scan_dir: Path,
- relative_path: str,
- payload: bytes,
- *,
- expected_root_identity: tuple[int, int] | None = None,
- ) -> None:
- path = scan_dir / relative_path
- path.parent.mkdir(parents=True, exist_ok=True)
- path.write_bytes(payload)
-
- def unlink_if_exists(scan_dir: Path, relative_path: str) -> None:
- (scan_dir / relative_path).unlink(missing_ok=True)
-
- backend.open_read_fd.side_effect = open_read_fd
- backend.atomic_write.side_effect = atomic_write
- backend.unlink_if_exists.side_effect = unlink_if_exists
+ backend = windows_file_backend()
with (
mock.patch.object(FINALIZER.os, "supports_dir_fd", set()),
@@ -1754,7 +1712,7 @@ The extraction root is not enforced.
self.write_json("findings.json", findings)
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "findings.json":
- artifact["sha256"] = FINALIZER._sha256_file(self.scan_dir / "findings.json")
+ artifact["sha256"] = self.sha256_file("findings.json")
self.write_json("scan-manifest.json", manifest)
with self.assertRaisesRegex(FINALIZER.ContractError, field):
@@ -2350,7 +2308,7 @@ The extraction root is not enforced.
manifest = self.read_json("scan-manifest.json")
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "coverage.json":
- artifact["sha256"] = FINALIZER._sha256_file(self.scan_dir / "coverage.json")
+ artifact["sha256"] = self.sha256_file("coverage.json")
elif artifact["path"] == receipt_ref:
artifact["path"] = legacy_ref
self.write_json("scan-manifest.json", manifest)
diff --git a/plugins/codex-security/tests/test_report_projection.py b/plugins/codex-security/tests/test_report_projection.py
index 3537d07fc..d797ce326 100644
--- a/plugins/codex-security/tests/test_report_projection.py
+++ b/plugins/codex-security/tests/test_report_projection.py
@@ -1,24 +1,9 @@
from __future__ import annotations
import copy
-import importlib.util
-from pathlib import Path
-from types import ModuleType
import pytest
-
-PLUGIN_DIR = Path(__file__).resolve().parent.parent
-
-
-def load_script(name: str) -> ModuleType:
- path = PLUGIN_DIR / "scripts" / f"{name}.py"
- spec = importlib.util.spec_from_file_location(name, path)
- if spec is None or spec.loader is None:
- raise RuntimeError(f"could not load {path}")
- module = importlib.util.module_from_spec(spec)
- spec.loader.exec_module(module)
- return module
-
+from workbench_test_support import load_script
PROJECTION = load_script("report_projection")
diff --git a/plugins/codex-security/tests/test_sarif_projection.py b/plugins/codex-security/tests/test_sarif_projection.py
index d1c369a6a..f7e151f5e 100644
--- a/plugins/codex-security/tests/test_sarif_projection.py
+++ b/plugins/codex-security/tests/test_sarif_projection.py
@@ -1,19 +1,15 @@
from __future__ import annotations
import copy
-import importlib.util
import json
import unittest
from pathlib import Path
+from workbench_test_support import load_script
+
PLUGIN_ROOT = Path(__file__).resolve().parent.parent
EXAMPLE_DIR = PLUGIN_ROOT / "examples" / "completed-scan"
-spec = importlib.util.spec_from_file_location(
- "sarif_finalizer", PLUGIN_ROOT / "scripts" / "finalize_scan_contract.py"
-)
-assert spec is not None and spec.loader is not None
-FINALIZER = importlib.util.module_from_spec(spec)
-spec.loader.exec_module(FINALIZER)
+FINALIZER = load_script("finalize_scan_contract", module_name="sarif_finalizer")
class SarifProjectionTest(unittest.TestCase):
diff --git a/plugins/codex-security/tests/test_validate_scan_contract.py b/plugins/codex-security/tests/test_validate_scan_contract.py
index a99993927..b637d3af4 100644
--- a/plugins/codex-security/tests/test_validate_scan_contract.py
+++ b/plugins/codex-security/tests/test_validate_scan_contract.py
@@ -1,50 +1,25 @@
from __future__ import annotations
-import importlib.util
import io
import json
-import shutil
import subprocess
import sys
-import tempfile
import unittest
from contextlib import redirect_stderr, redirect_stdout
from pathlib import Path
-from types import ModuleType
+
+from workbench_test_support import ScanFixtureTestCase, load_script
PLUGIN_ROOT = Path(__file__).resolve().parents[1]
EXAMPLE_SCAN = PLUGIN_ROOT / "examples" / "completed-scan"
-def load_validator() -> ModuleType:
- script = PLUGIN_ROOT / "scripts" / "validate_scan_contract.py"
- spec = importlib.util.spec_from_file_location("validate_scan_contract", script)
- if spec is None or spec.loader is None:
- raise RuntimeError(f"could not load {script}")
- module = importlib.util.module_from_spec(spec)
- spec.loader.exec_module(module)
- return module
-
-
-VALIDATOR = load_validator()
+VALIDATOR = load_script("validate_scan_contract")
-class ValidateScanContractTest(unittest.TestCase):
- def setUp(self) -> None:
- self.temp_dir = tempfile.TemporaryDirectory()
- self.scan_dir = Path(self.temp_dir.name) / "scan"
- shutil.copytree(EXAMPLE_SCAN, self.scan_dir)
- manifest = json.loads((self.scan_dir / "scan-manifest.json").read_text())
- findings = json.loads((self.scan_dir / "findings.json").read_text())
- coverage = json.loads((self.scan_dir / "coverage.json").read_text())
- report = VALIDATOR.FINALIZER._generate_report_projection(manifest, findings, coverage)
- (self.scan_dir / "report.md").write_bytes(report)
-
- def tearDown(self) -> None:
- self.temp_dir.cleanup()
-
- def read_json(self, name: str) -> dict[str, object]:
- return json.loads((self.scan_dir / name).read_text(encoding="utf-8"))
+class ValidateScanContractTest(ScanFixtureTestCase):
+ validator = VALIDATOR
+ example_scan = EXAMPLE_SCAN
def write_json(self, name: str, payload: object) -> None:
(self.scan_dir / name).write_text(json.dumps(payload, indent=2) + "\n")
@@ -71,9 +46,7 @@ class ValidateScanContractTest(unittest.TestCase):
for artifact in manifest["scan"]["artifacts"]
if artifact["path"] == "findings.json"
)
- findings_artifact["sha256"] = VALIDATOR.FINALIZER._sha256_file(
- self.scan_dir / "findings.json"
- )
+ findings_artifact["sha256"] = self.sha256_file("findings.json")
self.write_json("scan-manifest.json", manifest)
validated = VALIDATOR.validate_contract(self.scan_dir)
@@ -105,9 +78,7 @@ class ValidateScanContractTest(unittest.TestCase):
manifest = self.read_json("scan-manifest.json")
for artifact in manifest["scan"]["artifacts"]:
if artifact["path"] == "findings.json":
- artifact["sha256"] = VALIDATOR.FINALIZER._sha256_file(
- self.scan_dir / "findings.json"
- )
+ artifact["sha256"] = self.sha256_file("findings.json")
break
self.write_json("scan-manifest.json", manifest)
diff --git a/plugins/codex-security/tests/test_validate_tracking_source.py b/plugins/codex-security/tests/test_validate_tracking_source.py
index 5a725c2c1..0d83bdf35 100644
--- a/plugins/codex-security/tests/test_validate_tracking_source.py
+++ b/plugins/codex-security/tests/test_validate_tracking_source.py
@@ -1,53 +1,29 @@
from __future__ import annotations
import copy
-import importlib.util
+import hashlib
import io
import json
-import shutil
-import tempfile
import unittest
from contextlib import redirect_stderr, redirect_stdout
from pathlib import Path
-from types import ModuleType
+
+from workbench_test_support import ScanFixtureTestCase, load_script
PLUGIN_ROOT = Path(__file__).resolve().parents[1]
EXAMPLE_SCAN = PLUGIN_ROOT / "examples" / "completed-scan"
-def load_validator() -> ModuleType:
- script = PLUGIN_ROOT / "scripts" / "validate_tracking_source.py"
- spec = importlib.util.spec_from_file_location("validate_tracking_source", script)
- if spec is None or spec.loader is None:
- raise RuntimeError(f"could not load {script}")
- module = importlib.util.module_from_spec(spec)
- spec.loader.exec_module(module)
- return module
-
-
-VALIDATOR = load_validator()
+VALIDATOR = load_script("validate_tracking_source")
def test_completed_example_passes_tracking_validation() -> None:
assert VALIDATOR.validate_source(EXAMPLE_SCAN)
-class ValidateTrackingSourceTest(unittest.TestCase):
- def setUp(self) -> None:
- self.temp_dir = tempfile.TemporaryDirectory()
- self.scan_dir = Path(self.temp_dir.name) / "scan"
- shutil.copytree(EXAMPLE_SCAN, self.scan_dir)
- manifest = json.loads((self.scan_dir / "scan-manifest.json").read_text())
- findings = json.loads((self.scan_dir / "findings.json").read_text())
- coverage = json.loads((self.scan_dir / "coverage.json").read_text())
- report = VALIDATOR.FINALIZER._generate_report_projection(manifest, findings, coverage)
- (self.scan_dir / "report.md").write_bytes(report)
-
- def tearDown(self) -> None:
- self.temp_dir.cleanup()
-
- def read_json(self, name: str) -> dict[str, object]:
- return json.loads((self.scan_dir / name).read_text(encoding="utf-8"))
+class ValidateTrackingSourceTest(ScanFixtureTestCase):
+ validator = VALIDATOR
+ example_scan = EXAMPLE_SCAN
def write_json(self, name: str, payload: object) -> None:
(self.scan_dir / name).write_text(
@@ -64,7 +40,7 @@ class ValidateTrackingSourceTest(unittest.TestCase):
for item in manifest["scan"]["artifacts"]
if item["path"] == manifest["scan"]["findingsRef"]
)
- artifact["sha256"] = VALIDATOR.FINALIZER._sha256_bytes(findings_bytes)
+ artifact["sha256"] = hashlib.sha256(findings_bytes).hexdigest()
self.write_json("scan-manifest.json", manifest)
def rewrite_coverage_and_seal(self, coverage: dict[str, object]) -> None:
@@ -76,7 +52,7 @@ class ValidateTrackingSourceTest(unittest.TestCase):
for item in manifest["scan"]["artifacts"]
if item["path"] == manifest["scan"]["coverageRef"]
)
- artifact["sha256"] = VALIDATOR.FINALIZER._sha256_bytes(coverage_bytes)
+ artifact["sha256"] = hashlib.sha256(coverage_bytes).hexdigest()
self.write_json("scan-manifest.json", manifest)
def test_selects_the_finding_from_a_sealed_scan(self) -> None:
@@ -103,7 +79,7 @@ class ValidateTrackingSourceTest(unittest.TestCase):
for item in manifest["scan"]["artifacts"]
if item["path"] == manifest["scan"]["findingsRef"]
)
- artifact["sha256"] = VALIDATOR.FINALIZER._sha256_bytes(findings_bytes)
+ artifact["sha256"] = hashlib.sha256(findings_bytes).hexdigest()
self.write_json("scan-manifest.json", manifest)
with self.assertRaisesRegex(
diff --git a/plugins/codex-security/tests/test_windows_report_e2e.py b/plugins/codex-security/tests/test_windows_report_e2e.py
index 562f9ac66..fa52dab42 100644
--- a/plugins/codex-security/tests/test_windows_report_e2e.py
+++ b/plugins/codex-security/tests/test_windows_report_e2e.py
@@ -13,6 +13,7 @@ from workbench_test_support import (
SCRIPT,
create_saved_workspace,
start_delivered_scan,
+ windows_file_backend,
write_completed_contract,
)
@@ -44,27 +45,7 @@ def test_workbench_completion_and_exports_use_windows_file_backend(tmp_path: Pat
(scan_dir / "report.html").write_text("stale report")
namespace = runpy.run_path(str(SCRIPT), run_name="codex_security_workbench_db")
finalizer = sys.modules[namespace["finalize_scan"].__module__]
- backend = mock.Mock()
-
- def open_read_fd(root: Path, relative_path: str, _context: str) -> int:
- return os.open(root / relative_path, os.O_RDONLY)
-
- def atomic_write(
- root: Path,
- relative_path: str,
- payload: bytes,
- *,
- expected_root_identity: tuple[int, int] | None = None,
- ) -> None:
- path = root / relative_path
- path.parent.mkdir(parents=True, exist_ok=True)
- path.write_bytes(payload)
-
- backend.open_read_fd.side_effect = open_read_fd
- backend.atomic_write.side_effect = atomic_write
- backend.unlink_if_exists.side_effect = lambda root, relative_path: (
- root / relative_path
- ).unlink(missing_ok=True)
+ backend = windows_file_backend()
with (
mock.patch.object(finalizer.os, "supports_dir_fd", set()),
diff --git a/plugins/codex-security/tests/test_windows_scan_local_files.py b/plugins/codex-security/tests/test_windows_scan_local_files.py
index 454818433..e52ee9257 100644
--- a/plugins/codex-security/tests/test_windows_scan_local_files.py
+++ b/plugins/codex-security/tests/test_windows_scan_local_files.py
@@ -1,24 +1,12 @@
from __future__ import annotations
-import importlib.util
import os
from pathlib import Path
-from types import ModuleType
import pytest
+from workbench_test_support import load_script
-
-def load_windows_scan_local_files() -> ModuleType:
- script = Path(__file__).resolve().parent.parent / "scripts" / "windows_scan_local_files.py"
- spec = importlib.util.spec_from_file_location("windows_scan_local_files", script)
- if spec is None or spec.loader is None:
- raise RuntimeError(f"could not load {script}")
- module = importlib.util.module_from_spec(spec)
- spec.loader.exec_module(module)
- return module
-
-
-WINDOWS_FILES = load_windows_scan_local_files()
+WINDOWS_FILES = load_script("windows_scan_local_files")
@pytest.mark.parametrize(
diff --git a/plugins/codex-security/tests/test_windows_scope_streams.py b/plugins/codex-security/tests/test_windows_scope_streams.py
index b2157a41a..d6bc968c2 100644
--- a/plugins/codex-security/tests/test_windows_scope_streams.py
+++ b/plugins/codex-security/tests/test_windows_scope_streams.py
@@ -17,12 +17,13 @@ def test_scope_rejects_ntfs_streams_before_resolving_paths(monkeypatch, script:
path = Path(__file__).resolve().parents[1] / "scripts" / script
namespace = runpy.run_path(str(path), run_name="scope_stream_test")
resolve_scope = namespace["resolve_scope"]
- monkeypatch.setitem(resolve_scope.__globals__, "os", SimpleNamespace(name="nt"))
+ stream_component = namespace["windows_stream_component"]
+ monkeypatch.setitem(stream_component.__globals__, "os", SimpleNamespace(name="nt"))
monkeypatch.setitem(resolve_scope.__globals__, "Path", WindowsScopePath)
root = PureWindowsPath(r"C:\Repository")
error = namespace.get("InventoryError", SystemExit)
with pytest.raises(error, match="alternate data stream"):
resolve_scope(root, "src/app.py:stream")
assert namespace["windows_stream_component"](root / "src/app.py") is None
- monkeypatch.setitem(resolve_scope.__globals__, "os", SimpleNamespace(name="posix"))
+ monkeypatch.setitem(stream_component.__globals__, "os", SimpleNamespace(name="posix"))
assert namespace["windows_stream_component"](Path("src/app.py:fixture")) is None
diff --git a/plugins/codex-security/tests/test_workbench_db_exports.py b/plugins/codex-security/tests/test_workbench_db_exports.py
index 9ae58c681..eff80d588 100644
--- a/plugins/codex-security/tests/test_workbench_db_exports.py
+++ b/plugins/codex-security/tests/test_workbench_db_exports.py
@@ -1721,9 +1721,9 @@ def test_parent_draft_preserves_reconciled_candidate_identity_before_publication
with monkeypatch.context() as context:
context.setattr(results, "write_scan_local_bytes", interrupt_coverage)
with pytest.raises(OSError, match="Interrupted canonical publication"):
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
else:
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
head = json.loads((scan_dir / "checkpoint-head.json").read_text())
normalized_path = scan_dir / "checkpoints" / head["checkpoint"]
normalized_bytes = normalized_path.read_bytes()
@@ -1748,7 +1748,7 @@ def test_parent_draft_preserves_reconciled_candidate_identity_before_publication
raw["coverage"]["surfaces"] = documents["coverage"]["surfaces"]
raw_path.write_text(json.dumps(raw))
staged.write_text(json.dumps(documents))
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
assert normalized_path.read_bytes() == normalized_bytes
assert retained_raw.read_bytes() == raw_bytes
canonical = json.loads((scan_dir / "coverage.json").read_text())
@@ -1758,7 +1758,7 @@ def test_parent_draft_preserves_reconciled_candidate_identity_before_publication
raw["coverage"]["deferred"] = [pending]
raw_path.write_text(json.dumps(raw))
staged.write_text(json.dumps(documents))
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
head_bytes = (scan_dir / "checkpoint-head.json").read_bytes()
checkpoints = set((scan_dir / "checkpoints").iterdir())
# A rejected write must not leave a fresh closure for stopped recovery.
@@ -1769,7 +1769,7 @@ def test_parent_draft_preserves_reconciled_candidate_identity_before_publication
raw_path.write_text(json.dumps(raw))
args.expected_draft_digest = "0" * 64
with pytest.raises(SystemExit, match="scan_draft_conflict"):
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
assert (scan_dir / "checkpoint-head.json").read_bytes() == head_bytes
assert set((scan_dir / "checkpoints").iterdir()) == checkpoints
args.expected_draft_digest = None
@@ -1777,14 +1777,14 @@ def test_parent_draft_preserves_reconciled_candidate_identity_before_publication
documents["manifest"]["scan"]["target"] = None
staged.write_text(json.dumps(documents))
with pytest.raises(results.ContractError, match="target"):
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
assert (scan_dir / "checkpoint-head.json").read_bytes() == head_bytes
assert set((scan_dir / "checkpoints").iterdir()) == checkpoints
documents["manifest"]["scan"]["target"] = target
documents["coverage"]["deferred"].remove(pending)
documents["coverage"]["resolvedDeferred"] = raw["coverage"]["resolvedDeferred"]
staged.write_text(json.dumps(documents))
- db.write_scan_draft(connection, args)
+ results.write_scan_draft(db._WORKBENCH_DB_CONTEXT, connection, args)
canonical = json.loads((scan_dir / "coverage.json").read_text())
assert [row["id"] for row in canonical["deferred"]] == ["review-b"]
assert canonical["resolvedDeferred"] == raw["coverage"]["resolvedDeferred"]
diff --git a/plugins/codex-security/tests/test_workbench_timestamps.py b/plugins/codex-security/tests/test_workbench_timestamps.py
index e47650c64..ca2e8ecc4 100644
--- a/plugins/codex-security/tests/test_workbench_timestamps.py
+++ b/plugins/codex-security/tests/test_workbench_timestamps.py
@@ -5,6 +5,7 @@ import json
import sqlite3
from datetime import datetime, timezone
from pathlib import Path
+from types import SimpleNamespace
import pytest
@@ -52,7 +53,7 @@ def test_deep_scan_deadline_and_heartbeat_on_python310(monkeypatch, tmp_path: Pa
monkeypatch.syspath_prepend(str(Path(__file__).resolve().parents[1] / "scripts"))
deep = importlib.import_module("deep_scan_workbench")
monkeypatch.setattr(deep, "datetime", Python310DateTime)
- monkeypatch.setattr(deep, "now", lambda: "2026-08-15T12:00:00Z")
+ monkeypatch.setattr(deep, "_dependencies", SimpleNamespace(now=lambda: "2026-08-15T12:00:00Z"))
assert deep.deep_scan_deadline_reached(
{"created_at": "2026-08-15T11:00:00z", "max_time_hours": 1}
)
diff --git a/plugins/codex-security/tests/workbench_test_support.py b/plugins/codex-security/tests/workbench_test_support.py
index b0ec2919e..454783b07 100644
--- a/plugins/codex-security/tests/workbench_test_support.py
+++ b/plugins/codex-security/tests/workbench_test_support.py
@@ -1,22 +1,37 @@
from __future__ import annotations
import hashlib
+import importlib.util
import json
import os
import shlex
+import shutil
import sqlite3
import stat
import subprocess
import sys
+import tempfile
import uuid
from pathlib import Path
+from types import ModuleType
from typing import Any
+from unittest import TestCase, mock
SCRIPT = Path(__file__).resolve().parents[1] / "scripts" / "workbench_db.py"
SNAPSHOT_SCRIPT = Path(__file__).resolve().parents[1] / "scripts" / "snapshot_sqlite.py"
PLUGIN_MANIFEST = Path(__file__).resolve().parents[1] / ".codex-plugin" / "plugin.json"
+def load_script(name: str, *, module_name: str | None = None) -> ModuleType:
+ script = SCRIPT.parent / f"{name}.py"
+ spec = importlib.util.spec_from_file_location(module_name or name, script)
+ if spec is None or spec.loader is None:
+ raise RuntimeError(f"could not load {script}")
+ module = importlib.util.module_from_spec(spec)
+ spec.loader.exec_module(module)
+ return module
+
+
def source_plugin_version() -> str:
manifest = json.loads(PLUGIN_MANIFEST.read_text(encoding="utf-8"))
version = manifest.get("version")
@@ -446,3 +461,50 @@ def write_completed_contract(
(scan_dir / "coverage.json").write_text(json.dumps(coverage))
(scan_dir / "scan-manifest.json").write_text(json.dumps(manifest))
(scan_dir / "report.md").write_text("# Fixture report\n")
+
+
+def windows_file_backend() -> mock.Mock:
+ backend = mock.Mock()
+
+ def open_read_fd(scan_dir: Path, relative_path: str, _context: str) -> int:
+ return os.open(scan_dir / relative_path, os.O_RDONLY)
+
+ def atomic_write(
+ scan_dir: Path,
+ relative_path: str,
+ payload: bytes,
+ *,
+ expected_root_identity: tuple[int, int] | None = None,
+ ) -> None:
+ path = scan_dir / relative_path
+ path.parent.mkdir(parents=True, exist_ok=True)
+ path.write_bytes(payload)
+
+ def unlink_if_exists(scan_dir: Path, relative_path: str) -> None:
+ (scan_dir / relative_path).unlink(missing_ok=True)
+
+ backend.open_read_fd.side_effect = open_read_fd
+ backend.atomic_write.side_effect = atomic_write
+ backend.unlink_if_exists.side_effect = unlink_if_exists
+ return backend
+
+
+class ScanFixtureTestCase(TestCase):
+ def setUp(self) -> None:
+ self.temp_dir = tempfile.TemporaryDirectory()
+ self.scan_dir = Path(self.temp_dir.name) / "scan"
+ shutil.copytree(self.example_scan, self.scan_dir)
+ manifest = json.loads((self.scan_dir / "scan-manifest.json").read_text())
+ findings = json.loads((self.scan_dir / "findings.json").read_text())
+ coverage = json.loads((self.scan_dir / "coverage.json").read_text())
+ report = self.validator.FINALIZER._generate_report_projection(manifest, findings, coverage)
+ (self.scan_dir / "report.md").write_bytes(report)
+
+ def tearDown(self) -> None:
+ self.temp_dir.cleanup()
+
+ def read_json(self, name: str) -> dict[str, object]:
+ return json.loads((self.scan_dir / name).read_text(encoding="utf-8"))
+
+ def sha256_file(self, name: str) -> str:
+ return hashlib.sha256((self.scan_dir / name).read_bytes()).hexdigest()
diff --git a/sdk/typescript/tests-ts/deep-scan-timestamp-compat.test.ts b/sdk/typescript/tests-ts/deep-scan-timestamp-compat.test.ts
index f6e6629c9..a3810901d 100644
--- a/sdk/typescript/tests-ts/deep-scan-timestamp-compat.test.ts
+++ b/sdk/typescript/tests-ts/deep-scan-timestamp-compat.test.ts
@@ -19,6 +19,7 @@ if sys.version_info >= (3, 11):
deep_scan.datetime = Python310Datetime
case = json.loads(sys.argv[2])
deep_scan.now = lambda: case["now"]
+deep_scan.configure(deep_scan)
connection = sqlite3.connect(":memory:")
connection.execute("CREATE TABLE deep_scan_workers (scan_id TEXT, status TEXT)")
if case.get("activeWorker"):
diff --git a/sdk/typescript/tests-ts/deep-scan-workbench.test.ts b/sdk/typescript/tests-ts/deep-scan-workbench.test.ts
index 7da35e97a..aa23340c0 100644
--- a/sdk/typescript/tests-ts/deep-scan-workbench.test.ts
+++ b/sdk/typescript/tests-ts/deep-scan-workbench.test.ts
@@ -1,7 +1,7 @@
import { mkdir, readFile, writeFile } from "node:fs/promises";
import { dirname, join } from "node:path";
import { afterEach, describe, expect, test } from "bun:test";
-import { runWorkbench } from "../src/runtime.js";
+import { workbenchCommand as workbench } from "./support/workbench-command.js";
import { loadBundledRuntime, PLUGIN_ROOT } from "./plugin-root.js";
import { runPython } from "./support/python-probe.js";
import { createTemporaryDirectories } from "./support/temporary-directories.js";
@@ -36,6 +36,7 @@ const deepScanOwnershipProbe = [
"deep_scan.require_scan = lambda database, value: database.execute('SELECT * FROM scans WHERE id = ?', (value,)).fetchone()",
"deep_scan.require_workspace = lambda database, value: database.execute('SELECT * FROM workspaces WHERE id = ?', (value,)).fetchone()",
"deep_scan.now = lambda: 'after'",
+ "deep_scan.configure(deep_scan)",
"deep_scan.deep_scan_result = lambda database, value, *, start_disposition=None: {'startDisposition': start_disposition}",
"try:",
" result = deep_scan.begin_deep_scan_for_scan(connection, scan_id, 'requesting-thread', argparse.Namespace(claim_token=case['suppliedToken'], model=None, reasoning_effort=None))",
@@ -138,6 +139,7 @@ test("recovers an interrupted copied Deep Scan publication", async () => {
"connection.execute('INSERT INTO scans VALUES (?, ?)', ('scan', str(scan_dir)))",
"connection.execute('INSERT INTO deep_scan_workers VALUES (?, ?, ?, ?, ?)', ('scan', 'dedup', 'running', str(snapshot.parent.parent), 'now'))",
"deep_scan.require_scan = lambda database, value: database.execute('SELECT * FROM scans WHERE id = ?', (value,)).fetchone()",
+ "deep_scan.configure(deep_scan)",
"deep_scan.recover_candidate_ledger_publication(connection, 'scan')",
"print(json.dumps({'ledger': ledger.read_text(encoding='utf-8'), 'backup': backup.exists()}))",
].join("\n"),
@@ -169,20 +171,11 @@ test.each([
await writeFile(join(repository, "source.py"), "# synthetic source\n");
const python = Bun.which("python3") ?? Bun.which("python");
expect(python).not.toBeNull();
- const command = (args: string[], input?: string) =>
- runWorkbench(
- {
- python: python!,
- pluginRoot: PLUGIN_ROOT,
- environment: {
- ...process.env,
- CODEX_SECURITY_STATE_DIR: join(root, "state"),
- CODEX_HOME: join(root, "codex-home"),
- },
- },
- args,
- input,
- );
+ const command = workbench(python!, () => ({
+ ...process.env,
+ CODEX_SECURITY_STATE_DIR: join(root, "state"),
+ CODEX_HOME: join(root, "codex-home"),
+ }));
const registration = await command(
[
"register-cli-scan",
diff --git a/sdk/typescript/tests-ts/support/workbench-command.ts b/sdk/typescript/tests-ts/support/workbench-command.ts
new file mode 100644
index 000000000..b11051fc5
--- /dev/null
+++ b/sdk/typescript/tests-ts/support/workbench-command.ts
@@ -0,0 +1,19 @@
+import { runWorkbench } from "../../src/runtime.js";
+import { PLUGIN_ROOT } from "../plugin-root.js";
+
+export function workbenchCommand(
+ python: string,
+ environment: string | (() => NodeJS.ProcessEnv),
+) {
+ if (typeof environment === "string")
+ return workbenchCommand(python, () => ({
+ PATH: process.env["PATH"],
+ CODEX_SECURITY_STATE_DIR: environment,
+ }));
+ return (args: readonly string[], input?: string) =>
+ runWorkbench(
+ { python, pluginRoot: PLUGIN_ROOT, environment: environment() },
+ args,
+ input,
+ );
+}
diff --git a/sdk/typescript/tests-ts/workbench-canonical-paths.test.ts b/sdk/typescript/tests-ts/workbench-canonical-paths.test.ts
index c15134c6d..bfd9ab755 100644
--- a/sdk/typescript/tests-ts/workbench-canonical-paths.test.ts
+++ b/sdk/typescript/tests-ts/workbench-canonical-paths.test.ts
@@ -35,6 +35,7 @@ const simulatedPathProbe = [
"deep_scan.Path = SimulatedPath",
"deep_scan.os = SimpleNamespace(path=path_module)",
"deep_scan.require_canonical_scan_directory = lambda path: path",
+ "deep_scan.configure(deep_scan)",
"try:",
" result = deep_scan.deep_scan_path({'scan_dir': root}, supplied, 'Worker prompt path', kind='file')",
"except SystemExit:",
@@ -65,6 +66,7 @@ const realFilesystemProbe = [
" artifact_name = 'prompts/prompt.txt'",
" candidate_name = 'prompt.txt'",
"deep_scan.require_canonical_scan_directory = workbench.require_canonical_scan_directory",
+ "deep_scan.configure(deep_scan)",
"def accepted(action):",
" try:",
" action()",