@atmem/architect 0.0.0-stage → 0.2.0

package/src/runtime/EnterpriseDocument.java

@@ -0,0 +1,89 @@
+/* Credentialless, single-document Tika bridge. Configuration is operator-owned.
+ * Run inside a network-isolated, resource-limited container; Tika is not a sandbox.
+ * No URL, filename, parser option or environment setting comes from a document.
+ */
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.charset.StandardCharsets;
+import java.util.Base64;
+import java.util.Set;
+import org.apache.tika.detect.DefaultDetector;
+import org.apache.tika.io.TikaInputStream;
+import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
+import org.apache.tika.parser.AutoDetectParser;
+import org.apache.tika.parser.ParseContext;
+import org.apache.tika.parser.RecursiveParserWrapper;
+import org.apache.tika.parser.microsoft.OfficeParserConfig;
+import org.apache.tika.parser.ocr.TesseractOCRConfig;
+import org.apache.tika.sax.BasicContentHandlerFactory;
+import org.apache.tika.sax.RecursiveParserWrapperHandler;
+
+public final class EnterpriseDocument {
+    private static final int LIMIT = 200000;
+    private static final Set<String> TYPES = Set.of(
+        "text/plain", "text/html", "application/rtf",
+        "application/msword", "application/vnd.ms-excel", "application/vnd.ms-powerpoint",
+        "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+        "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+        "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+        "application/vnd.oasis.opendocument.text", "application/vnd.oasis.opendocument.spreadsheet",
+        "application/vnd.oasis.opendocument.presentation");
+
+    public static void main(String[] args) {
+        try {
+            if (args.length != 1) throw new IllegalArgumentException();
+            Path input = Path.of(args[0]);
+            if (!Files.isRegularFile(input) || Files.size(input) > 10000000 || Files.size(input) == 0)
+                throw new IllegalArgumentException();
+            Metadata root = new Metadata();
+            String detected;
+            try (TikaInputStream stream = TikaInputStream.get(input)) {
+                detected = new DefaultDetector().detect(stream, root).toString();
+            }
+            if (!TYPES.contains(detected)) {
+                System.out.print("rejected\nunsupported_format\n");
+                System.exit(2);
+            }
+            AutoDetectParser parser = new AutoDetectParser();
+            ParseContext context = new ParseContext();
+            OfficeParserConfig office = new OfficeParserConfig();
+            office.setExtractMacros(false);
+            office.setIncludeDeletedContent(false);
+            office.setIncludeMoveFromContent(false);
+            office.setIncludeHeadersAndFooters(true);
+            office.setIncludeSlideNotes(true);
+            context.set(OfficeParserConfig.class, office);
+            TesseractOCRConfig ocr = new TesseractOCRConfig();
+            ocr.setLanguage("eng");
+            ocr.setTimeoutSeconds(15);
+            ocr.setEnableImagePreprocessing(false);
+            context.set(TesseractOCRConfig.class, ocr);
+            RecursiveParserWrapperHandler handler = new RecursiveParserWrapperHandler(
+                new BasicContentHandlerFactory(BasicContentHandlerFactory.HANDLER_TYPE.TEXT, LIMIT), 64);
+            // Recursive metadata exposes caught embedded/parse exceptions and write
+            // limits, which plain CLI text output can otherwise quietly conceal.
+            try (TikaInputStream stream = TikaInputStream.get(input)) {
+                new RecursiveParserWrapper(parser).parse(stream, handler, root, context);
+            }
+            StringBuilder text = new StringBuilder();
+            for (Metadata item : handler.getMetadataList()) {
+                for (String name : item.names()) {
+                    if (name.startsWith("X-TIKA:EXCEPTION") || name.startsWith("X-TIKA:WARN") ||
+                        name.contains("write_limit_reached") || name.contains("embedded_resource_limit_reached"))
+                        throw new IllegalArgumentException();
+                }
+                String content = item.get(TikaCoreProperties.TIKA_CONTENT);
+                if (content != null && !content.isBlank()) text.append(content.strip()).append('\n');
+                if (text.length() > LIMIT) throw new IllegalArgumentException();
+            }
+            byte[] result = text.toString().strip().getBytes(StandardCharsets.UTF_8);
+            if (result.length == 0 || result.length > LIMIT) throw new IllegalArgumentException();
+            System.out.print("ok\n" + detected + "\n" + Base64.getEncoder().encodeToString(result) + "\n");
+        } catch (Throwable failure) {
+            // Native parser messages can contain private text, filenames and URLs.
+            System.out.print("rejected\nextraction_failed\n");
+            System.exit(2);
+        }
+    }
+}

package/src/runtime/approved_http_destination.py

@@ -0,0 +1,72 @@
+"""A concrete, explicit normalized external-action HTTP contract.
+
+This is not an invented vendor API. Destination operators must verify this exact
+idempotent contract or implement a reviewed native mapping. A model/visitor cannot
+choose the endpoint, credentials, identity header, method or replay policy.
+"""
+import hashlib
+import http.client
+import json
+import re
+import ssl
+from urllib.parse import urlsplit
+from durable_workflow import canonical,identifier
+from workflow_errors import ExternalOutcomeUnknown,DestinationNotAttempted,InvalidDispatch
+from workflow_contract import normalized_contract,bind_payload_contract,bind_application_contract,bind_review_policy
+
+class ApprovedHttpDestination:
+    def __init__(self,endpoint,credential,*,destination_scope,idempotency_verified=False,tls_context=None,allowed_kinds=('document',),payload_contract=None,application_contract_sha256=None,review_policy=None):
+        try:target=urlsplit(endpoint)
+        except (TypeError,ValueError):raise ValueError('Invalid operator destination')from None
+        if (not isinstance(endpoint,str)or len(endpoint)>2000 or target.scheme!='https' or not target.hostname or target.username or target.password
+            or target.query or target.fragment or any(ord(c)<33 for c in endpoint)):raise ValueError('Use a fixed reviewed HTTPS destination without embedded credentials, query or fragment')
+        try:port=target.port or 443
+        except ValueError:raise ValueError('Invalid destination port')from None
+        if not 1<=port<=65535 or not callable(credential)or type(idempotency_verified)is not bool:raise ValueError('Invalid owned destination configuration')
+        if not isinstance(allowed_kinds,(tuple,list,set,frozenset))or not allowed_kinds or not set(allowed_kinds)<={'document','draft','account_change'}:raise ValueError('Invalid owned operation allowlist')
+        context=tls_context or ssl.create_default_context()
+        if not isinstance(context,ssl.SSLContext)or context.verify_mode!=ssl.CERT_REQUIRED or not context.check_hostname:raise ValueError('TLS certificate and hostname verification are mandatory')
+        self.host=target.hostname;self.port=port;self.path=target.path or'/';self.credential=credential;self.idempotency_verified=idempotency_verified;self.context=context;self.allowed_kinds=frozenset(allowed_kinds)
+        self.contract=bind_payload_contract(normalized_contract(endpoint,destination_scope),payload_contract)
+        if application_contract_sha256 is not None:self.contract=bind_application_contract(self.contract,application_contract_sha256)
+        self.contract=bind_review_policy(self.contract,payload_contract,review_policy)
+        self.review_policy=review_policy
+        self.payload_contract=payload_contract
+    def __call__(self,row,*,idempotency_key,payload_sha256):
+        try:identifier(idempotency_key)
+        except (ValueError,TypeError):raise InvalidDispatch('Invalid approved operation identity')from None
+        if (str(row.get('id'))!=idempotency_key or row.get('state')!='dispatching' or not row.get('approved_by')or row.get('approved_hash')!=payload_sha256
+            or row.get('kind')not in self.allowed_kinds or not isinstance(payload_sha256,str)or not re.fullmatch(r'[0-9a-f]{64}',payload_sha256)):raise InvalidDispatch('A current allowed approved dispatch snapshot is required')
+        if row.get('destination_contract')!=self.contract:raise InvalidDispatch('Approved integration identity no longer matches this destination')
+        if self.payload_contract is not None:
+            try:
+                if self.contract['payloadContractSha256']!=self.payload_contract.sha256:raise ValueError()
+                self.payload_contract.validate('source',row['source_ref']);self.payload_contract.validate('result',row['result'])
+                if self.review_policy:self.review_policy.validate_result(row['source_ref'],row['result'])
+            except (KeyError,ValueError):raise InvalidDispatch('Approved application contract no longer matches this destination')from None
+        snapshot={'kind':row['kind'],'source':row['source_ref'],'destination':row['destination_contract'],'result':row['result'],'revision':row['result_revision']}
+        if hashlib.sha256(canonical(snapshot,140000)).hexdigest()!=payload_sha256:raise InvalidDispatch('Approved payload snapshot changed')
+        # Source storage references and credentials stay inside our boundary.
+        body=canonical({'operation_id':idempotency_key,'payload_sha256':payload_sha256,'kind':row['kind'],'result':row['result']},70000)
+        try:token=self.credential()
+        except Exception:raise DestinationNotAttempted('Destination credential unavailable; no request made')from None
+        if not isinstance(token,str)or not 1<=len(token)<=16000 or any(ord(c)<33 or ord(c)>126 for c in token):raise DestinationNotAttempted('Destination credential unavailable; no request made')
+        connection=http.client.HTTPSConnection(self.host,self.port,timeout=8,context=self.context)
+        try:
+            connection.request('POST',self.path,body=body,headers={'Content-Type':'application/json','Accept':'application/json','Authorization':'Bearer '+token,'Idempotency-Key':idempotency_key})
+            response=connection.getresponse();raw=response.read(8193)
+            # No redirects, inferred success, 202-as-completed, or error-body logs.
+            if response.status not in {200,201}or len(raw)>8192 or response.getheader('Content-Type','').split(';')[0].lower()!='application/json':raise ExternalOutcomeUnknown('Destination completion is unconfirmed')
+            def unique(pairs):
+                result={}
+                for key,value in pairs:
+                    if key in result:raise ValueError('Duplicate receipt field')
+                    result[key]=value
+                return result
+            reply=json.loads(raw,object_pairs_hook=unique)
+            if (not isinstance(reply,dict)or set(reply)!={'operation_id','payload_sha256','external_id','status'}or reply['operation_id']!=idempotency_key
+                or reply['payload_sha256']!=payload_sha256 or reply['status']!='completed' or not isinstance(reply['external_id'],str)or not re.fullmatch(r'[A-Za-z0-9_.:/-]{1,200}',reply['external_id'])):raise ExternalOutcomeUnknown('Destination completion is unconfirmed')
+            return {'externalId':reply['external_id'],'operationId':idempotency_key,'payloadSha256':payload_sha256}
+        except ExternalOutcomeUnknown:raise
+        except Exception:raise ExternalOutcomeUnknown('Destination completion is unconfirmed')from None
+        finally:connection.close()

package/src/runtime/aws-lambda-qualification-lock.json

@@ -0,0 +1,7 @@
+{
+  "image": "public.ecr.aws/lambda/python:3.13@sha256:181a54632ba2cb7945ac633bdac72c13d7b0e461a32931966ae93b429a3436b9",
+  "platform": "linux/amd64",
+  "reviewedAt": "2026-11-10",
+  "source": "https://docs.aws.amazon.com/lambda/latest/dg/python-image.html",
+  "boundary": "Official AWS Python 3.13 x86_64 image manifest observed from public ECR. Local qualification only; Lambda managed-runtime patch level and account behavior require separate destination evidence."
+}

package/src/runtime/aws-python-lock.json

@@ -0,0 +1,82 @@
+{
+  "requirements": "boto3==1.43.123 --hash=sha256:d18d66e1e13faa0c25d0b1b8c5ce946e29fb5fc0b6456b655a93a678ff4cc313\nbotocore==1.43.123 --hash=sha256:e30580f9c1282c321529238ff8ff75066096a2adbd0a293c8ae592c21d1733a3\ns3transfer==0.19.3 --hash=sha256:24eb7535804306355e31c952a90022bf5b2f9d760f9a11062c7493a217c08553\njmespath==1.1.0 --hash=sha256:a5663118de4908c91729bea0acadca56526eb2698e83de10cd116ae0f4e97c64\npython-dateutil==2.9.0.post0 --hash=sha256:a8b2bc7bffae282281c8140a97d3aa9c14da0b136dfe83f850eea9a5f7470427\nsix==1.17.0 --hash=sha256:4721f391ed90541fddacab5acf947aa0d3dc7d27b2e1e8eda2be8970586c3274\nurllib3==2.8.0 --hash=sha256:0cf3cae568d36aa9576b28dfb35f11328f1cb974ca7647d9475ebb86c75ac6e3\n",
+  "sources": [
+    {
+      "package": "boto3",
+      "version": "1.43.123",
+      "source": "https://pypi.org/pypi/boto3/1.43.123/json",
+      "wheels": [
+        {
+          "filename": "boto3-1.43.123-py3-none-any.whl",
+          "sha256": "d18d66e1e13faa0c25d0b1b8c5ce946e29fb5fc0b6456b655a93a678ff4cc313"
+        }
+      ]
+    },
+    {
+      "package": "botocore",
+      "version": "1.43.123",
+      "source": "https://pypi.org/pypi/botocore/1.43.123/json",
+      "wheels": [
+        {
+          "filename": "botocore-1.43.123-py3-none-any.whl",
+          "sha256": "e30580f9c1282c321529238ff8ff75066096a2adbd0a293c8ae592c21d1733a3"
+        }
+      ]
+    },
+    {
+      "package": "s3transfer",
+      "version": "0.19.3",
+      "source": "https://pypi.org/pypi/s3transfer/0.19.3/json",
+      "wheels": [
+        {
+          "filename": "s3transfer-0.19.3-py3-none-any.whl",
+          "sha256": "24eb7535804306355e31c952a90022bf5b2f9d760f9a11062c7493a217c08553"
+        }
+      ]
+    },
+    {
+      "package": "jmespath",
+      "version": "1.1.0",
+      "source": "https://pypi.org/pypi/jmespath/1.1.0/json",
+      "wheels": [
+        {
+          "filename": "jmespath-1.1.0-py3-none-any.whl",
+          "sha256": "a5663118de4908c91729bea0acadca56526eb2698e83de10cd116ae0f4e97c64"
+        }
+      ]
+    },
+    {
+      "package": "python-dateutil",
+      "version": "2.9.0.post0",
+      "source": "https://pypi.org/pypi/python-dateutil/2.9.0.post0/json",
+      "wheels": [
+        {
+          "filename": "python_dateutil-2.9.0.post0-py2.py3-none-any.whl",
+          "sha256": "a8b2bc7bffae282281c8140a97d3aa9c14da0b136dfe83f850eea9a5f7470427"
+        }
+      ]
+    },
+    {
+      "package": "six",
+      "version": "1.17.0",
+      "source": "https://pypi.org/pypi/six/1.17.0/json",
+      "wheels": [
+        {
+          "filename": "six-1.17.0-py2.py3-none-any.whl",
+          "sha256": "4721f391ed90541fddacab5acf947aa0d3dc7d27b2e1e8eda2be8970586c3274"
+        }
+      ]
+    },
+    {
+      "package": "urllib3",
+      "version": "2.8.0",
+      "source": "https://pypi.org/pypi/urllib3/2.8.0/json",
+      "wheels": [
+        {
+          "filename": "urllib3-2.8.0-py3-none-any.whl",
+          "sha256": "0cf3cae568d36aa9576b28dfb35f11328f1cb974ca7647d9475ebb86c75ac6e3"
+        }
+      ]
+    }
+  ]
+}