agentme 0.22.0 → 0.23.0

package/.xdrs/agentme/edrs/application/021-ai-workflow-development-standards.md

@@ -134,6 +134,23 @@ Names MUST NOT use generic labels such as `node1`, `process`, or `run`. Each nam
 
 Judge nodes use a **prefix** convention instead of a suffix: the name MUST start with `evaluate_` followed by the subject being judged (e.g. `evaluate_progress`, `evaluate_quality`, `evaluate_completeness`, `evaluate_relevance`). This makes judge nodes immediately distinguishable from all other node types at a glance.
 
+**Grouping prefix for related nodes:** When multiple nodes deal with the same subject, entity, or workflow region, SHOULD use a shared grouping word as a prefix followed by a verb and the role suffix. The pattern is `<group>_<verb>_<role_suffix>`. This makes the graph topology scannable and clusters related nodes together alphabetically in logs, traces, and code.
+
+```python
+# Nodes grouped under the "invoice" subject
+def invoice_fetch_tool(state): ...       # fetches invoice data from an API
+def invoice_validate_step(state): ...    # validates invoice fields deterministically
+def invoice_summarize_llm(state): ...    # summarizes invoice content with an LLM
+def invoice_review_agent(state): ...     # runs an agent loop to review the invoice
+
+graph.add_node("invoice_fetch_tool", invoice_fetch_tool)
+graph.add_node("invoice_validate_step", invoice_validate_step)
+graph.add_node("invoice_summarize_llm", invoice_summarize_llm)
+graph.add_node("invoice_review_agent", invoice_review_agent)
+```
+
+The grouping prefix is optional for workflows where all nodes clearly belong to a single domain. It MUST be used when a workflow spans multiple subjects or regions (e.g. `invoice_*`, `payment_*`, `notification_*`) to prevent name collisions and to make the graph structure self-documenting.
+
 #### 10-workflow-unit-testing
 
 All LLM calls within workflow nodes are external API calls and MUST be mocked in unit tests per [agentme-edr-018](018-ai-llm-development-standards.md) rule `04-unit-test-mocking`. Workflow unit tests must run fully offline with no real LLM provider calls.

package/.xdrs/agentme/edrs/application/028-ai-eval-standards.md

@@ -182,6 +182,12 @@ Where $\hat{p}$ is observed accuracy and $n$ is sample count. Accuracy and F1 ar
 - MLflow run: experiment `workflow-document-review/eval-basic` — view with `mlflow ui`
 ```
 
+#### 04-eval-mlflow-unique-port
+
+Each `evals/<component>/eval-<name>/Makefile` MUST start its MLflow tracking server on a **unique port** to prevent conflicts when multiple eval Makefiles are run concurrently or in parallel (e.g., in CI or across multiple terminal sessions).
+
+Ports MUST be statically assigned per eval scenario and MUST NOT reuse the default `5000` port (reserved for `dev-mlflow` per [agentme-edr-008](../devops/008-common-targets.md) rule `09-ai-project-dev-targets`). Assign ports starting at `5100` and incrementing by 1 for each additional eval scenario across the entire project.
+
 ## References
 
 - [agentme-edr-007](../principles/007-project-quality-standards.md) — Project quality standards: when evals are required per AI tier (rule `09-ai-project-testing-requirements`) and statistical model eval targets (rule `07-statistical-models-must-have-eval-targets`)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentme",
-  "version": "0.22.0",
+  "version": "0.23.0",
   "description": "",
   "dependencies": {
     "filedist": "^0.36.0"