Merge branch 'main' into feature/upload-multiple-files-at-once-to-bytes

boefjes/.ci/docker-compose.yml

@@ -8,6 +8,7 @@ services:
     command: sh -c 'python -m pytest -v tests/integration'
     depends_on:
       - ci_katalogus-db
+      - ci_katalogus
     env_file:
       - .ci/.env.test
     volumes:
@@ -103,8 +104,15 @@ services:
         hard: 262144
 
   ci_katalogus:
-    image: "docker.io/wiremock/wiremock:2.34.0"
-    volumes:
-      - .ci/wiremock:/home/wiremock
+    build:
+      context: ..
+      dockerfile: boefjes/Dockerfile
+      args:
+        - ENVIRONMENT=dev
+    command: uvicorn boefjes.katalogus.root:app --host 0.0.0.0 --port 8080
+    depends_on:
+      - ci_katalogus-db
     env_file:
       - .ci/.env.test
+    volumes:
+      - .:/app/boefjes

boefjes/boefjes/api.py

@@ -151,8 +151,7 @@ def get_task(task_id, scheduler_client):
 def create_boefje_meta(task, local_repository):
     boefje = task.data.boefje
     boefje_resource = local_repository.by_id(boefje.id)
-    env_keys = boefje_resource.environment_keys
-    environment = get_environment_settings(task.data, env_keys) if env_keys else {}
+    environment = get_environment_settings(task.data, boefje_resource.schema)
 
     organization = task.data.organization
     input_ooi = task.data.input_ooi

boefjes/boefjes/dependencies/plugins.py

@@ -98,7 +98,7 @@ def clone_settings_to_organisation(self, from_organisation: str, to_organisation
             self.set_enabled_by_id(plugin_id, to_organisation, enabled=True)
 
     def upsert_settings(self, settings: dict, organisation_id: str, plugin_id: str):
-        self._assert_settings_match_schema(settings, organisation_id, plugin_id)
+        self._assert_settings_match_schema(settings, plugin_id)
         self._put_boefje(plugin_id)
 
         return self.config_storage.upsert(organisation_id, plugin_id, settings=settings)
@@ -122,14 +122,14 @@ def _put_boefje(self, boefje_id: str) -> None:
 
         try:
             self.plugin_storage.boefje_by_id(boefje_id)
-        except PluginNotFound:
+        except PluginNotFound as e:
             try:
                 plugin = self.local_repo.by_id(boefje_id)
             except KeyError:
-                raise
+                raise e
 
             if plugin.type != "boefje":
-                raise
+                raise e
             self.plugin_storage.create_boefje(plugin)
 
     def _put_normalizer(self, normalizer_id: str) -> None:
@@ -150,12 +150,7 @@ def _put_normalizer(self, normalizer_id: str) -> None:
     def delete_settings(self, organisation_id: str, plugin_id: str):
         self.config_storage.delete(organisation_id, plugin_id)
 
-        try:
-            self._assert_settings_match_schema({}, organisation_id, plugin_id)
-        except SettingsNotConformingToSchema:
-            logger.warning("Making sure %s is disabled for %s because settings are deleted", plugin_id, organisation_id)
-
-            self.set_enabled_by_id(plugin_id, organisation_id, False)
+        # We don't check the schema anymore because we can provide entries through the global environment as well
 
     def schema(self, plugin_id: str) -> dict | None:
         try:
@@ -184,9 +179,7 @@ def description(self, plugin_id: str, organisation_id: str) -> str:
             return ""
 
     def set_enabled_by_id(self, plugin_id: str, organisation_id: str, enabled: bool):
-        if enabled:
-            all_settings = self.get_all_settings(organisation_id, plugin_id)
-            self._assert_settings_match_schema(all_settings, organisation_id, plugin_id)
+        # We don't check the schema anymore because we can provide entries through the global environment as well
 
         try:
             self._put_boefje(plugin_id)
@@ -195,14 +188,14 @@ def set_enabled_by_id(self, plugin_id: str, organisation_id: str, enabled: bool)
 
         self.config_storage.upsert(organisation_id, plugin_id, enabled=enabled)
 
-    def _assert_settings_match_schema(self, all_settings: dict, organisation_id: str, plugin_id: str):
+    def _assert_settings_match_schema(self, all_settings: dict, plugin_id: str):
         schema = self.schema(plugin_id)
 
         if schema:  # No schema means that there is nothing to assert
             try:
                 validate(instance=all_settings, schema=schema)
             except ValidationError as e:
-                raise SettingsNotConformingToSchema(organisation_id, plugin_id, e.message) from e
+                raise SettingsNotConformingToSchema(plugin_id, e.message) from e
 
     def _set_plugin_enabled(self, plugin: PluginType, organisation_id: str) -> PluginType:
         with contextlib.suppress(KeyError, NotFound):

boefjes/boefjes/job_handler.py

@@ -7,6 +7,8 @@
 import httpx
 import structlog
 from httpx import HTTPError
+from jsonschema.exceptions import ValidationError
+from jsonschema.validators import validate
 
 from boefjes.clients.bytes_client import BytesAPIClient
 from boefjes.config import settings
@@ -15,6 +17,7 @@
 from boefjes.local_repository import LocalPluginRepository
 from boefjes.plugins.models import _default_mime_types
 from boefjes.runtime_interfaces import BoefjeJobRunner, Handler, NormalizerJobRunner
+from boefjes.storage.interfaces import SettingsNotConformingToSchema
 from octopoes.api.models import Affirmation, Declaration, Observation
 from octopoes.connector.octopoes import OctopoesAPIConnector
 from octopoes.models import Reference, ScanLevel
@@ -35,30 +38,42 @@ def get_octopoes_api_connector(org_code: str) -> OctopoesAPIConnector:
     return OctopoesAPIConnector(str(settings.octopoes_api), org_code)
 
 
-def get_environment_settings(boefje_meta: BoefjeMeta, environment_keys: list[str]) -> dict[str, str]:
+def get_environment_settings(boefje_meta: BoefjeMeta, schema: dict | None = None) -> dict[str, str]:
     try:
         katalogus_api = str(settings.katalogus_api).rstrip("/")
         response = httpx.get(
             f"{katalogus_api}/v1/organisations/{boefje_meta.organization}/{boefje_meta.boefje.id}/settings",
             timeout=30,
         )
         response.raise_for_status()
-        environment = response.json()
-
-        # Add prefixed BOEFJE_* global environment variables
-        for key, value in os.environ.items():
-            if key.startswith("BOEFJE_"):
-                katalogus_key = key.split("BOEFJE_", 1)[1]
-                # Only pass the environment variable if it is not explicitly set through the katalogus,
-                # if and only if they are defined in boefje.json
-                if katalogus_key in environment_keys and katalogus_key not in environment:
-                    environment[katalogus_key] = value
-
-        return {k: str(v) for k, v in environment.items() if k in environment_keys}
     except HTTPError:
         logger.exception("Error getting environment settings")
         raise
 
+    allowed_keys = schema.get("properties", []) if schema else []
+    new_env = {
+        key.split("BOEFJE_", 1)[1]: value
+        for key, value in os.environ.items()
+        if key.startswith("BOEFJE_") and key in allowed_keys
+    }
+
+    settings_from_katalogus = response.json()
+
+    for key, value in settings_from_katalogus.items():
+        if key in allowed_keys:
+            new_env[key] = value
+
+    # The schema, besides dictating that a boefje cannot run if it is not matched, also provides an extra safeguard:
+    # it is possible to inject code if arguments are passed that "escape" the call to a tool. Hence, we should enforce
+    # the schema somewhere and make the schema as strict as possible.
+    if schema is not None:
+        try:
+            validate(instance=new_env, schema=schema)
+        except ValidationError as e:
+            raise SettingsNotConformingToSchema(boefje_meta.boefje.id, e.message) from e
+
+    return new_env
+
 
 class BoefjeHandler(Handler):
     def __init__(
@@ -97,10 +112,8 @@ def handle(self, boefje_meta: BoefjeMeta) -> None:
 
             boefje_meta.arguments["input"] = ooi.serialize()
 
-        env_keys = boefje_resource.environment_keys
-
         boefje_meta.runnable_hash = boefje_resource.runnable_hash
-        boefje_meta.environment = get_environment_settings(boefje_meta, env_keys) if env_keys else {}
+        boefje_meta.environment = get_environment_settings(boefje_meta, boefje_resource.schema)
 
         mime_types = _default_mime_types(boefje_meta.boefje)
 

boefjes/boefjes/katalogus/plugins.py

@@ -123,7 +123,6 @@ class BoefjeIn(BaseModel):
     version: str | None = None
     created: datetime.datetime | None = None
     description: str | None = None
-    environment_keys: list[str] = Field(default_factory=list)
     scan_level: int = 1
     consumes: set[str] = Field(default_factory=set)
     produces: set[str] = Field(default_factory=set)
@@ -167,7 +166,6 @@ class NormalizerIn(BaseModel):
     version: str | None = None
     created: datetime.datetime | None = None
     description: str | None = None
-    environment_keys: list[str] = Field(default_factory=list)
     consumes: list[str] = Field(default_factory=list)  # mime types (and/ or boefjes)
     produces: list[str] = Field(default_factory=list)  # oois
 

boefjes/boefjes/migrations/versions/870fc302b852_remove_environment_keys_field.py

@@ -0,0 +1,37 @@
+"""Remove environment keys field
+
+Revision ID: 870fc302b852
+Revises: 5be152459a7b
+Create Date: 2024-08-20 06:08:20.943924
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "870fc302b852"
+down_revision = "5be152459a7b"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column("boefje", "environment_keys")
+    op.drop_column("normalizer", "environment_keys")
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column(
+        "normalizer",
+        sa.Column("environment_keys", postgresql.ARRAY(sa.VARCHAR(length=128)), autoincrement=False, nullable=False),
+    )
+    op.add_column(
+        "boefje",
+        sa.Column("environment_keys", postgresql.ARRAY(sa.VARCHAR(length=128)), autoincrement=False, nullable=False),
+    )
+    # ### end Alembic commands ###

boefjes/boefjes/migrations/versions/f9de6eb7824b_introduce_boefjeconfig_model.py

@@ -95,7 +95,7 @@ def upgrade() -> None:
             str(boefje.scan_level),
             list(boefje.consumes),
             list(boefje.produces),
-            boefje.environment_keys,
+            ["TEST_KEY"],
             boefje.oci_image,
             boefje.oci_arguments,
             boefje.version,
@@ -137,7 +137,7 @@ def upgrade() -> None:
             str(boefje.scan_level),
             list(boefje.consumes),
             list(boefje.produces),
-            boefje.environment_keys,
+            ["TEST_KEY"],
             boefje.oci_image,
             boefje.oci_arguments,
             boefje.version,
@@ -177,7 +177,7 @@ def upgrade() -> None:
             normalizer.description,
             normalizer.consumes,
             normalizer.produces,
-            normalizer.environment_keys,
+            ["TEST_KEY"],
             normalizer.version,
         )
         for normalizer in normalizers_to_insert

boefjes/boefjes/models.py

@@ -17,7 +17,6 @@ class Plugin(BaseModel):
     version: str | None = None
     created: datetime.datetime | None = None
     description: str | None = None
-    environment_keys: list[str] = Field(default_factory=list)
     enabled: bool = False
     static: bool = True  # We need to differentiate between local and remote plugins to know which ones can be deleted
 

boefjes/boefjes/plugins/kat_adr_finding_types/boefje.json

@@ -1,7 +1,7 @@
 {
   "id": "adr-finding-types",
   "name": "ADR Finding Types",
-  "description": "Hydrate information of ADR finding types",
+  "description": "Hydrate information on API Design Rules (ADR) finding types for common design mistakes.",
   "consumes": [
     "ADRFindingType"
   ],

boefjes/boefjes/plugins/kat_adr_finding_types/normalizer.json

@@ -1,5 +1,7 @@
 {
   "id": "kat_adr_finding_types_normalize",
+  "name": "API Design Rules (ADR) Finding Types",
+  "description": "Parse API Design Rules (ADR) finding types.",
   "consumes": [
     "boefje/adr-finding-types"
   ],

boefjes/boefjes/plugins/kat_adr_validator/boefje.json

@@ -1,7 +1,7 @@
 {
   "id": "adr-validator",
   "name": "API Design Rules validator",
-  "description": "Validate if an API conforms to the API Design Rules",
+  "description": "Validate if an API conforms to the API Design Rules (ADR).",
   "consumes": [
     "RESTAPI"
   ],

boefjes/boefjes/plugins/kat_adr_validator/normalizer.json

@@ -1,5 +1,7 @@
 {
   "id": "adr-validator-normalize",
+  "name": "API Design Rules validator",
+  "description": "TODO",
   "consumes": [
     "boefje/adr-validator"
   ],

boefjes/boefjes/plugins/kat_answer_parser/normalizer.json

@@ -1,5 +1,7 @@
 {
   "id": "kat_answer_parser",
+  "name": "Answer Parser",
+  "description": "Parses the answers from Config objects.",
   "consumes": [
     "answer"
   ],

boefjes/boefjes/plugins/kat_binaryedge/boefje.json

@@ -1,13 +1,10 @@
 {
   "id": "binaryedge",
   "name": "BinaryEdge",
-  "description": "Use BinaryEdge to find open ports with vulnerabilities that are found on that port",
+  "description": "Use BinaryEdge to find open ports with vulnerabilities. Requires a BinaryEdge API key.",
   "consumes": [
     "IPAddressV4",
     "IPAddressV6"
   ],
-  "environment_keys": [
-    "BINARYEDGE_API"
-  ],
   "scan_level": 2
 }

boefjes/boefjes/plugins/kat_binaryedge/containers/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_containers",
+  "name": "BinaryEdge containers",
   "consumes": [
     "boefje/binaryedge"
   ],

boefjes/boefjes/plugins/kat_binaryedge/databases/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_databases",
+  "name": "BinaryEdge databases",
   "consumes": [
     "boefje/binaryedge"
   ],

boefjes/boefjes/plugins/kat_binaryedge/http_web/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_http_web",
+  "name": "BinaryEdge Websites",
   "consumes": [
     "boefje/binaryedge"
   ],

boefjes/boefjes/plugins/kat_binaryedge/message_queues/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_message_queues",
+  "name": "BinaryEdge message queues",
   "consumes": [
     "boefje/binaryedge"
   ],

boefjes/boefjes/plugins/kat_binaryedge/protocols/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_protocols",
+  "name": "BinaryEdge protocols",
   "consumes": [
     "boefje/binaryedge"
   ],

boefjes/boefjes/plugins/kat_binaryedge/remote_desktop/normalizer.json

@@ -1,5 +1,6 @@
 {
   "id": "kat_binaryedge_remote_desktop",
+  "name": "Binary Edge remote desktop",
   "consumes": [
     "boefje/binaryedge"
   ],