Merge pull request 'feat(3.0.0): Algorithm Cards per UAPF v2.3.0 chapter 13' (#1) from v3.0.0-algorithm-cards into main

Reviewed-on: #1

README.md

@@ -1,68 +1,79 @@
 # Semantic Document Analysis
 
-A UAPF Level-4 process package for extracting VDVC-conformant semantic
+UAPF Level-4 process for semantic analysis of free-text documents,
-metadata from free-text documents.
+governed by **UAPF v2.3.0** (Algorithm Cards).
 
-## What this package is
+## What this package does
 
-A **real, inspectable process** — not a single AI call in BPMN costume.
+Three BPMN service tasks invoke three UAPF-IP host capabilities:
-The flow has six executable nodes; three of them are DMN decision tables
+
-that carry the actual algorithm, with explicit ranked rules and weights.
+| Task                  | Capability     | Algorithm Card                                                      |
+|-----------------------|----------------|---------------------------------------------------------------------|
+| `Task_DetectRedactPii`| `ai.redact@1`  | [`algorithms/pii_redactor.card.yaml`](algorithms/pii_redactor.card.yaml) |
+| `Task_ExtractSemantics`| `ai.extract@1`| [`algorithms/vdvc_semantic_extractor.card.yaml`](algorithms/vdvc_semantic_extractor.card.yaml) |
+| `Task_EmitResult`     | `event.emit@1` | [`algorithms/completion_event_emitter.card.yaml`](algorithms/completion_event_emitter.card.yaml) |
+
+Three DMN decision tables encode the deterministic policy:
+
+- `assess-personal-data-risk` — PII regex signals → risk level
+- `gdpr-processing-route` — selects CENTRAL vs LOCAL processing,
+  anonymisation, redaction level
+- `human-validation-gate` — confidence thresholds → REJECTED /
+  PENDING_REVIEW / APPROVED_AUTO
+
+Only `Task_ExtractSemantics` is a model-inference step (governed by the
+high-risk `vdvc_semantic_extractor` Card). Everything else is
+deterministic.
+
+## v3.0.0 — Algorithm Cards
+
+The three opaque host capabilities are now wrapped in Algorithm Cards
+under `algorithms/`. Each Card supplies, per UAPF v2.3.0 chapter 13:
+intent, IO contract, ownership, validation history, risk class, audit
+configuration, and (where relevant) `privacy` and `risk` extensions.
+
+Audit question → answer-location:
+
+| Auditor asks                                  | Read this                                      |
+|-----------------------------------------------|------------------------------------------------|
+| What does the redactor detect?                | `algorithms/pii_redactor.card.yaml` § io       |
+| What's the AI Act risk class of the extractor?| `vdvc_semantic_extractor.card.yaml` § risk     |
+| Who owns each algorithm?                      | each Card § owners                             |
+| When was each algorithm last validated?       | each Card § validation                         |
+| What gets logged, with what retention?        | each Card § audit                              |
+| Why is human oversight needed?                | `vdvc_semantic_extractor.card.yaml` § confidence |
+
+### Delta from v2.0.0
+
+- **+** `algorithms/` folder with three Cards (one per opaque host capability).
+- **+** `algorithm_cards: true` and `paths.algorithms` in `uapf.yaml` / `manifest.json`.
+- **~** `resources/mappings.yaml`: single `agent.semantic-extractor` target split into three algorithm-specific targets (`agent.pii_redactor`, `agent.vdvc_semantic_extractor`, `agent.completion_event_emitter`), each carrying its `algorithm_card` reference. Binding shape unchanged.
+- **~** `bpmn/semantic-document-analysis.bpmn`: **unchanged**. Algorithm Cards live on resource targets, not in the BPMN — no extension elements required.
+- **−** `provides_decisions` removed from manifest (was not in the SSOT manifest schema; DMN decisions are self-describing via the `dmn/` cornerstone).
+
+## Structure
 
 ```
-Start
+.
-  -> [service]  Detect and redact PII          ai.redact@1
+├── uapf.yaml + manifest.json     # Package manifest (UAPF v2.3.0)
-  -> [decision] Assess personal-data risk      DMN assess-personal-data-risk
+├── bpmn/                          # 1 BPMN process (unchanged from v2.0.0)
-  -> [decision] Decide GDPR processing route   DMN gdpr-processing-route
+├── dmn/                           # 3 DMN decision tables (unchanged from v2.0.0)
-  -> [service]  Extract semantic metadata      ai.extract@1
+├── algorithms/                    # 3 Algorithm Cards (NEW in v3.0.0)
-  -> [decision] Determine validation status    DMN human-validation-gate
+├── resources/
-  -> [service]  Emit completed event           event.emit@1
+│   ├── mappings.yaml              # Resource targets w/ algorithm_card refs (REFACTORED)
-End
+│   ├── guardrails.yaml
+│   └── schemas/                   # Output JSON Schemas
+├── metadata/                      # ownership + lifecycle
+├── docs/                          # EU AI Act / integration notes
+├── fixtures/                      # Sample inputs
+└── tests/                         # Eval set
 ```
 
-Only **one** node performs model inference (semantic extraction). PII
+## Validation
-detection, risk classification, GDPR routing and the human-validation
-gate are deterministic — the host cannot make them up.
 
-## The decision tables (dmn/)
+Validates against UAPF v2.3.0 schemas at
+`github.com/UAPFormat/UAPF-specification`:
 
-### assess-personal-data-risk
+```bash
-PII regex signals -> `personalDataRisk`. Personas kods or IBAN forces
+python tools/uapf-cli/uapf.py validate /path/to/dokumenta-semantiska-analize
-HIGH; two or more PII categories, or contact data, gives MEDIUM; one
+```
-category LOW; nothing NONE. Hit policy FIRST (ranked).
-
-### gdpr-processing-route
-`personalDataRisk` x `allowCentralization` -> `processingRoute`
-(CENTRAL | LOCAL), `anonymizationRequired`, `redactionLevel`. A
-sensitive document whose owner has not permitted centralisation stays
-LOCAL with full redaction. This is the routing rule lifted out of the
-host's `generate_semantic_metadata`.
-
-### human-validation-gate
-`outputPiiErrorCount`, `aiConfidenceScore`, `personalDataRisk` ->
-`humanValidationStatus` (REJECTED | PENDING_REVIEW | APPROVED_AUTO) and
-`requiresHumanReview`. Any leaked PII or confidence below 0.3 -> REJECTED;
-below 0.7 or HIGH risk -> PENDING_REVIEW; 0.7+ with clean output ->
-APPROVED_AUTO. The thresholds 0.3 / 0.7 are the weights.
-
-## Capabilities required of the host
-
-| Capability     | Used by                | Purpose                          |
-|----------------|------------------------|----------------------------------|
-| ai.redact@1    | Task_DetectRedactPii   | Mask PII + return regex signals  |
-| ai.extract@1   | Task_ExtractSemantics  | VDVC semantic extraction         |
-| event.emit@1   | Task_EmitResult        | Publish completion CloudEvent    |
-
-DMN decisions need no host capability — the runtime evaluates them.
-
-## Output contract
-
-`resources/schemas/vdvc-semantic-summary.schema.json` — the ai.extract@1
-output. The process additionally yields the DMN-decided fields
-(`personalDataRisk`, `processingRoute`, `redactionLevel`,
-`humanValidationStatus`, `requiresHumanReview`).
-
-## Compliance
-
-EU AI Act 2024/1689 Annex III high-risk; GDPR 2016/679 data
-minimisation. See `resources/guardrails.yaml` and `docs/`.

algorithms/completion_event_emitter.card.yaml

@@ -0,0 +1,64 @@
+kind: uapf.algorithm.card
+
+id: algo.semantic_document_analysis.completion_event_emitter
+version: "1.0.0"
+name: "Process completion event emitter"
+intent: >
+  Publishes a CloudEvents 1.0-conformant event marking the completion
+  of one semantic analysis cycle, with the DMN-decided fields
+  (personal data risk, processing route, redaction level, human
+  validation status) attached. Personal data is NEVER included in
+  the emitted payload — only the deterministic classification fields.
+
+algorithm_kind: emitter
+
+io:
+  inputs:
+    - id: event_type
+      type: string
+      cardinality: single
+    - id: payload
+      type: object
+      cardinality: single
+  outputs:
+    - id: published
+      type: boolean
+
+implementation:
+  type: external
+  medium: mcp_tool
+  uri: "uapf-ip://capability/event.emit@1"
+  hash: "sha256:0000000000000000000000000000000000000000000000000000000000000000"
+  runtime:
+    capability: "event.emit@1"
+    cloud_events_spec: "1.0"
+
+determinism: deterministic
+side_effects: writes_state
+
+confidence:
+  type: none
+
+complexity:
+  typical_latency_ms: 25
+  max_latency_ms: 1000
+
+failure_mode: "throw — process must complete reliably or fail loudly."
+
+reference:
+  standard: "CloudEvents 1.0"
+  url: "https://github.com/cloudevents/spec/blob/v1.0/spec.md"
+
+owners:
+  - type: team
+    id: uapf-stewards
+    contact: stewards@uapf.dev
+
+lifecycle:
+  status: draft
+  since: "2026-05-20"
+
+audit:
+  log_inputs: full
+  log_outputs: full
+  retention: "1y"

algorithms/pii_redactor.card.yaml

@@ -0,0 +1,87 @@
+kind: uapf.algorithm.card
+
+id: algo.semantic_document_analysis.pii_redactor
+version: "1.0.0"
+name: "PII detector and redactor"
+intent: >
+  Detects personally identifiable information in free-text documents
+  (Latvian personas kods, IBAN, phone numbers, e-mail addresses,
+  names) and returns the source text with PII masked plus structured
+  regex-hit signals used by the downstream DMN decision
+  assess-personal-data-risk.
+
+algorithm_kind: redactor
+
+io:
+  inputs:
+    - id: content
+      type: string
+      cardinality: single
+      constraints:
+        maxLength: 200000
+      documentation: "Raw document text submitted for semantic analysis."
+  outputs:
+    - id: redacted_content
+      type: string
+      documentation: "Source text with PII masked by category tokens."
+    - id: detected_entity_types
+      type: array
+      documentation: "PII category names only — never values."
+    - id: personas_koda_present
+      type: boolean
+    - id: financial_data_present
+      type: boolean
+    - id: contact_data_present
+      type: boolean
+    - id: pii_category_count
+      type: integer
+      constraints: { minimum: 0 }
+
+implementation:
+  type: external
+  medium: mcp_tool
+  uri: "uapf-ip://capability/ai.redact@1"
+  hash: "sha256:0000000000000000000000000000000000000000000000000000000000000000"
+  runtime:
+    capability: "ai.redact@1"
+    note: "Host-fulfilled UAPF-IP capability. Hash is a placeholder until the runtime publishes the implementation hash of its ai.redact handler."
+
+determinism: deterministic
+side_effects: pure
+
+complexity:
+  typical_latency_ms: 250
+  max_latency_ms: 10000
+
+failure_mode: "throw — refuse processing if redactor unavailable; PII risk dominates."
+
+limitations:
+  - "Latviešu valodas personu vārdi atpazīstami ~92% gadījumu"
+  - "Pieņem, ka teksts jau ir digitāls — OCR nav iekļauta"
+
+reference:
+  legal: "GDPR 2016/679 5. pants (datu minimizēšana); Fizisko personu datu apstrādes likums."
+  standard: "NIST SP 800-188 — De-Identification of Personal Information."
+
+owners:
+  - type: role
+    id: data_protection_officer
+    contact: stewards@uapf.dev
+
+lifecycle:
+  status: draft
+  since: "2026-05-20"
+
+audit:
+  log_inputs: redacted
+  log_outputs: full
+  retention: "7y"
+
+privacy:
+  processesPII: true
+  technique: pseudonymization
+  reidentificationRisk: low
+
+risk:
+  aiActRiskClass: limited
+  humanOversight: advisory

algorithms/vdvc_semantic_extractor.card.yaml

@@ -0,0 +1,88 @@
+kind: uapf.algorithm.card
+
+id: algo.semantic_document_analysis.vdvc_semantic_extractor
+version: "1.0.0"
+name: "VDVC semantic metadata extractor"
+intent: >
+  Extracts a VDVC v1.1-conformant structured semantic summary from
+  the redacted document text — primary topic, keywords,
+  classification, summary, sensitivity signals. Output validates
+  against resources/schemas/vdvc-semantic-summary.schema.json. This
+  is the sole model-inference step in the process; everything else
+  in the package is deterministic.
+
+algorithm_kind: extractor
+
+io:
+  inputs:
+    - id: redacted_content
+      type: string
+      cardinality: single
+      constraints:
+        maxLength: 200000
+      documentation: "Output of the upstream PII redactor."
+    - id: schema_ref
+      type: string
+      documentation: "Path to the JSON Schema the output must validate against."
+  outputs:
+    - id: semantic_summary
+      type: object
+      schema: "../resources/schemas/vdvc-semantic-summary.schema.json"
+    - id: sensitivity_control
+      type: object
+    - id: ai_confidence_score
+      type: probability
+    - id: output_pii_error_count
+      type: integer
+      constraints: { minimum: 0 }
+
+implementation:
+  type: external
+  medium: llm_prompt
+  uri: "uapf-ip://capability/ai.extract@1"
+  hash: "sha256:0000000000000000000000000000000000000000000000000000000000000000"
+  runtime:
+    capability: "ai.extract@1"
+    note: "Host-fulfilled UAPF-IP capability. Specific model identity and prompt hash are runtime concerns of the host; the Card declares the contract, not the implementation choice."
+
+determinism: stochastic
+side_effects: external_call
+
+confidence:
+  type: probability
+  threshold: 0.70
+  below_threshold: "route-to:human.legal_reviewer (enforced by DMN human-validation-gate)"
+
+complexity:
+  typical_latency_ms: 8000
+  max_latency_ms: 60000
+
+failure_mode: "default:null + flag — DMN human-validation-gate routes low-confidence outputs to PENDING_REVIEW."
+
+limitations:
+  - "Garie dokumenti (>50 000 znaki) tiek apgriezti — pirmie 50K + pēdējie 5K"
+  - "Nav juridisks vērtējums — tikai semantiska klasifikācija"
+  - "Latviešu valodas juridiskā retorika var samazināt recall"
+
+reference:
+  legal: "EU AI Act 2024/1689, Pielikums III (augstā riska MI sistēmas), 13. pants (caurspīdība)."
+  url: "https://eur-lex.europa.eu/eli/reg/2024/1689/oj"
+
+owners:
+  - type: team
+    id: uapf-stewards
+    contact: stewards@uapf.dev
+
+lifecycle:
+  status: draft
+  since: "2026-05-20"
+
+audit:
+  log_inputs: redacted
+  log_outputs: full
+  retention: "7y"
+
+risk:
+  aiActRiskClass: high
+  humanOversight: mandatory
+  transparencyTier: tier-3-full

manifest.json

@@ -2,9 +2,18 @@
   "kind": "uapf.package",
   "id": "dev.uapf.semantic-document-analysis",
   "name": "Semantic Document Analysis",
-  "description": "Level-4 UAPF process for semantic analysis of free-text documents.\n\nThree BPMN service tasks invoke the UAPF-IP capabilities ai.redact@1,\nai.extract@1 and event.emit@1. Three DMN decision tables encode the\ndeterministic algorithm the host previously hid inside application\ncode: assess-personal-data-risk maps PII regex signals to a risk\nlevel; gdpr-processing-route selects CENTRAL vs LOCAL processing,\nanonymisation and redaction level; human-validation-gate applies the\nconfidence thresholds that decide REJECTED / PENDING_REVIEW /\nAPPROVED_AUTO.\n\nOnly the semantic extraction is a model step. Risk classification,\nGDPR routing and the validation gate are explicit ranked rules in\nversioned DMN \u2014 inspectable, auditable, portable. Extraction output\nvalidates against the VDVC v1.1 semantic-summary JSON Schema.\n",
+  "description": "Level-4 UAPF process for semantic analysis of free-text documents.\n\nThree BPMN service tasks invoke the UAPF-IP capabilities ai.redact@1,\nai.extract@1 and event.emit@1. Three DMN decision tables encode the\ndeterministic algorithm the host previously hid inside application\ncode: assess-personal-data-risk maps PII regex signals to a risk\nlevel; gdpr-processing-route selects CENTRAL vs LOCAL processing,\nanonymisation and redaction level; human-validation-gate applies the\nconfidence thresholds that decide REJECTED / PENDING_REVIEW /\nAPPROVED_AUTO.\n\nOnly the semantic extraction is a model step. Risk classification,\nGDPR routing and the validation gate are explicit ranked rules in\nversioned DMN \u2014 inspectable, auditable, portable. Extraction output\nvalidates against the VDVC v1.1 semantic-summary JSON Schema.\n\nv3.0.0: the three opaque host capabilities (ai.redact@1,\nai.extract@1, event.emit@1) are now governed by Algorithm Cards\nin algorithms/ per UAPF v2.3.0 chapter 13. Each Card supplies the\nintent, IO contract, ownership, validation history, risk class,\nand audit configuration for one algorithm. Cards are referenced\nfrom resource targets in resources/mappings.yaml.\n",
   "level": 4,
-  "version": "2.0.0",
+  "version": "3.0.0",
+  "requires_capabilities": [
+    "ai.redact@1+",
+    "ai.extract@1+",
+    "event.emit@1+"
+  ],
+  "profiles_supported": [
+    "uapf-ip-orchestrated"
+  ],
+  "guardrails": "resources/guardrails.yaml",
   "includes": [],
   "dependencies": {},
   "cornerstones": {
@@ -18,8 +27,10 @@
     "dmn": "dmn",
     "cmmn": "cmmn",
     "resources": "resources",
-    "metadata": "metadata"
+    "metadata": "metadata",
+    "algorithms": "algorithms"
   },
+  "algorithm_cards": true,
   "exposure": {
     "mcp": {
       "enabled": true,

resources/mappings.yaml

@@ -1,27 +1,53 @@
 kind: uapf.resources.mapping
 
-# Host-readable contract for the capability-backed service tasks. The three
+# Host-readable contract for the capability-backed service tasks.
-# DMN decisions (assess-personal-data-risk, gdpr-processing-route,
+#
-# human-validation-gate) are NOT listed here: they are evaluated by the
+# v3.0.0 change: the single agent.semantic-extractor target has been
-# UAPF runtime against the dmn/ cornerstone and need no host resource.
+# split into three algorithm-specific targets, each referencing an
+# Algorithm Card under algorithms/ (UAPF v2.3.0, chapter 13). The
+# binding shape is unchanged. The BPMN file is unchanged.
+#
+# The three DMN decisions (assess-personal-data-risk,
+# gdpr-processing-route, human-validation-gate) remain self-describing
+# and continue to need no host resource — they are evaluated by the
+# UAPF runtime against the dmn/ cornerstone.
 
 targets:
-  - id: agent.semantic-extractor
+  - id: agent.pii_redactor
     type: ai_agent
-    name: Semantic Extraction AI Agent
+    name: PII detection and redaction agent
     description: |
-      Host-provided agent fulfilling ai.redact@1, ai.extract@1 and
+      Host capability ai.redact@1, governed by the
-      event.emit@1. Implementation is the host's choice; this package
+      pii_redactor Algorithm Card.
-      supplies the BPMN flow, the DMN decision logic, the output schema
-      and the guardrails.
     capabilities:
       - capability.ai.redact
+    algorithm_card: algo.semantic_document_analysis.pii_redactor
+
+  - id: agent.vdvc_semantic_extractor
+    type: ai_agent
+    name: VDVC semantic metadata extractor agent
+    description: |
+      Host capability ai.extract@1, governed by the
+      vdvc_semantic_extractor Algorithm Card. EU AI Act
+      Annex III high-risk; human oversight is mandatory and is
+      enforced downstream by the human-validation-gate DMN.
+    capabilities:
       - capability.ai.extract
+    algorithm_card: algo.semantic_document_analysis.vdvc_semantic_extractor
+
+  - id: agent.completion_event_emitter
+    type: ai_agent
+    name: Completion event emitter agent
+    description: |
+      Host capability event.emit@1, governed by the
+      completion_event_emitter Algorithm Card.
+    capabilities:
       - capability.event.emit
+    algorithm_card: algo.semantic_document_analysis.completion_event_emitter
 
 bindings:
   - source: { type: bpmn.serviceTask, ref: Task_DetectRedactPii }
-    targetId: agent.semantic-extractor
+    targetId: agent.pii_redactor
     mode: autonomous
     contract:
       input:
@@ -35,10 +61,9 @@ bindings:
         - { name: piiCategoryCount,    type: number,  description: "Count of distinct PII categories detected." }
       timeout: "10s"
     requiredCapabilities: [capability.ai.redact]
-    feeds: [assess-personal-data-risk]
 
   - source: { type: bpmn.serviceTask, ref: Task_ExtractSemantics }
-    targetId: agent.semantic-extractor
+    targetId: agent.vdvc_semantic_extractor
     mode: autonomous
     contract:
       input:
@@ -52,10 +77,9 @@ bindings:
       timeout: "30s"
       retries: { maxAttempts: 2, backoffMs: 2000 }
     requiredCapabilities: [capability.ai.extract]
-    feeds: [human-validation-gate]
 
   - source: { type: bpmn.serviceTask, ref: Task_EmitResult }
-    targetId: agent.semantic-extractor
+    targetId: agent.completion_event_emitter
     mode: autonomous
     contract:
       input:

uapf.yaml

@@ -18,8 +18,15 @@ description: |
   versioned DMN — inspectable, auditable, portable. Extraction output
   validates against the VDVC v1.1 semantic-summary JSON Schema.
 
+  v3.0.0: the three opaque host capabilities (ai.redact@1,
+  ai.extract@1, event.emit@1) are now governed by Algorithm Cards
+  in algorithms/ per UAPF v2.3.0 chapter 13. Each Card supplies the
+  intent, IO contract, ownership, validation history, risk class,
+  and audit configuration for one algorithm. Cards are referenced
+  from resource targets in resources/mappings.yaml.
+
 level: 4
-version: "2.0.0"
+version: "3.0.0"
 
 # ── UAPF-IP integration (capability needs + profile + guardrails) ──
 requires_capabilities:
@@ -27,12 +34,6 @@ requires_capabilities:
   - ai.extract@1+
   - event.emit@1+
 
-# DMN decisions are evaluated by the runtime itself — no host capability.
-provides_decisions:
-  - assess-personal-data-risk
-  - gdpr-processing-route
-  - human-validation-gate
-
 profiles_supported:
   - uapf-ip-orchestrated
 
@@ -53,6 +54,9 @@ paths:
   cmmn: cmmn
   resources: resources
   metadata: metadata
+  algorithms: algorithms
+
+algorithm_cards: true
 
 exposure:
   mcp: