Shorten publication run identifiers

.github/scripts/resolve_run_context.py

@@ -46,7 +46,6 @@ def _github_actions_run_id(env: Mapping[str, str]) -> str:
     return build_run_id(
         github_run_id=env.get("GITHUB_RUN_ID", ""),
         github_run_attempt=env.get("GITHUB_RUN_ATTEMPT", "1"),
-        github_sha=env.get("GITHUB_SHA", ""),
     )
 
 

.github/workflows/local_area_promote.yaml

@@ -4,7 +4,7 @@ on:
   workflow_dispatch:
     inputs:
       run_id:
-        description: 'Run ID to promote (e.g. usdata-gha123456-a1-abcdef12)'
+        description: 'Run ID to promote (e.g. usdata-gha123456-a1)'
         required: true
         type: string
 
@@ -22,7 +22,7 @@ jobs:
       MODAL_TOKEN_ID: ${{ secrets.MODAL_TOKEN_ID }}
       MODAL_TOKEN_SECRET: ${{ secrets.MODAL_TOKEN_SECRET }}
       MODAL_ENVIRONMENT: main
-      US_DATA_MODAL_APP_PREFIX: policyengine-us-data-pub
+      US_DATA_MODAL_APP_PREFIX: us-data
       US_DATA_RUN_ID: ${{ github.event.inputs.run_id }}
 
     steps:

.github/workflows/pipeline.yaml

@@ -77,7 +77,7 @@ jobs:
     runs-on: ubuntu-latest
     env:
       MODAL_ENVIRONMENT: main
-      US_DATA_MODAL_APP_PREFIX: policyengine-us-data-pub
+      US_DATA_MODAL_APP_PREFIX: us-data
       US_DATA_RUN_ID: ${{ inputs.run_id || '' }}
     steps:
       - uses: actions/checkout@v6

.github/workflows/push.yaml

@@ -20,7 +20,7 @@ jobs:
       github_run_url: ${{ steps.run-context.outputs.github_run_url }}
     env:
       MODAL_ENVIRONMENT: main
-      US_DATA_MODAL_APP_PREFIX: policyengine-us-data-pub
+      US_DATA_MODAL_APP_PREFIX: us-data
     steps:
       - uses: actions/checkout@v4
       - uses: actions/setup-python@v5

changelog.d/981.changed.md

@@ -0,0 +1 @@
+Shortened publication run IDs and default Modal app names while preserving candidate-scoped staging identity.

modal_app/pipeline.py

@@ -827,6 +827,28 @@ def _write_validation_diagnostics(
 # ── Orchestrator ─────────────────────────────────────────────────
 
 
+def _new_run_metadata(
+    *,
+    run_id: str,
+    branch: str,
+    sha: str,
+    candidate_version: str,
+    release_version: str,
+    run_context: RunContext,
+) -> RunMetadata:
+    return RunMetadata(
+        run_id=run_id,
+        branch=branch,
+        sha=sha,
+        version=candidate_version,
+        candidate_version=candidate_version,
+        release_version=release_version,
+        start_time=datetime.now(timezone.utc).isoformat(),
+        status="running",
+        **_metadata_run_fields(run_context),
+    )
+
+
 @app.function(
     image=image,
     cpu=2,
@@ -1021,18 +1043,13 @@ def run_pipeline(
             )
         _apply_run_context_env(current_run_context)
         run_id = current_run_context.run_id
-        meta = RunMetadata(
+        meta = _new_run_metadata(
             run_id=run_id,
             branch=branch,
             sha=sha,
-            version=candidate_version,
             candidate_version=candidate_version,
             release_version=release_version,
-            base_release_version=current_run_context.base_release_version,
-            release_bump=current_run_context.release_bump,
-            start_time=datetime.now(timezone.utc).isoformat(),
-            status="running",
-            **_metadata_run_fields(current_run_context),
+            run_context=current_run_context,
         )
 
     # Create run directory

policyengine_us_data/utils/run_context.py

@@ -25,7 +25,7 @@
 DATA_PACKAGE_VERSION_ENV = "US_DATA_PACKAGE_VERSION"
 MODAL_APP_NAME_ENV = "US_DATA_MODAL_APP_NAME"
 MODAL_ENVIRONMENT_ENV = "US_DATA_MODAL_ENVIRONMENT"
-DEFAULT_MODAL_APP_PREFIX = "policyengine-us-data-pub"
+DEFAULT_MODAL_APP_PREFIX = "us-data"
 DEFAULT_MODAL_ENVIRONMENT = "main"
 DEFAULT_MAX_RESOURCE_NAME_LENGTH = 64
 VALID_RELEASE_BUMPS = frozenset({"major", "minor", "patch"})
@@ -109,14 +109,12 @@ def build_run_id(
     *,
     github_run_id: str,
     github_run_attempt: str,
-    github_sha: str,
 ) -> str:
     """Build a deterministic run ID from GitHub Actions identity."""
     if not github_run_id:
         raise ValueError("github_run_id is required")
     attempt = github_run_attempt or "1"
-    sha = (github_sha or "unknown")[:8]
-    return sanitize_run_id(f"usdata-gha{github_run_id}-a{attempt}-{sha}")
+    return sanitize_run_id(f"usdata-gha{github_run_id}-a{attempt}")
 
 
 def build_modal_resource_name(
@@ -132,22 +130,48 @@ def build_modal_resource_name(
     )
 
 
-def staging_prefix(
-    run_id: str = "",
+def candidate_run_segment(
+    run_id: str,
     candidate_version: str = "",
     *,
     version: str = "",
 ) -> str:
-    if not run_id:
-        return "staging"
+    """Return the shared candidate/run segment for staging and app names."""
     resolved_run_id = sanitize_run_id(run_id)
     resolved_candidate_version = candidate_version or version
     if not resolved_candidate_version:
-        return f"staging/{resolved_run_id}"
-    staging_scope = sanitize_staging_version(
+        return resolved_run_id
+    return sanitize_staging_version(
         f"{sanitize_staging_version(resolved_candidate_version)}-{resolved_run_id}"
     )
-    return f"staging/{staging_scope}"
+
+
+def build_modal_app_name(
+    run_id: str,
+    candidate_version: str = "",
+    *,
+    prefix: str = DEFAULT_MODAL_APP_PREFIX,
+    max_length: int = DEFAULT_MAX_RESOURCE_NAME_LENGTH,
+) -> str:
+    """Build a safe Modal app name from candidate scope and run ID."""
+    return build_modal_resource_name(
+        candidate_run_segment(run_id, candidate_version),
+        prefix=prefix,
+        max_length=max_length,
+    )
+
+
+def staging_prefix(
+    run_id: str = "",
+    candidate_version: str = "",
+    *,
+    version: str = "",
+) -> str:
+    if not run_id:
+        return "staging"
+    return (
+        f"staging/{candidate_run_segment(run_id, candidate_version, version=version)}"
+    )
 
 
 def github_run_url(env: Mapping[str, str]) -> str:
@@ -368,8 +392,9 @@ def from_env(
             or env.get(MODAL_APP_NAME_ENV, "")
             or env.get("MODAL_APP_NAME", "")
             or (
-                build_modal_resource_name(
+                build_modal_app_name(
                     resolved_run_id,
+                    candidate_version=resolved_candidate_version,
                     prefix=modal_app_prefix,
                 )
                 if resolved_run_id

tests/unit/calibration/test_chunked_matrix_modal.py

@@ -70,13 +70,13 @@ def from_name(app_name: str, function_name: str, **kwargs):
             return object()
 
     monkeypatch.setitem(sys.modules, "modal", SimpleNamespace(Function=_FakeFunction))
-    monkeypatch.setenv("US_DATA_MODAL_APP_NAME", "policyengine-us-data-pub-run")
+    monkeypatch.setenv("US_DATA_MODAL_APP_NAME", "us-data-run")
     monkeypatch.setenv("US_DATA_MODAL_ENVIRONMENT", "main")
 
     _lookup_worker_function()
 
     assert captured == {
-        "app_name": "policyengine-us-data-pub-run",
+        "app_name": "us-data-run",
         "function_name": "build_matrix_chunk_worker",
         "kwargs": {"environment_name": "main"},
     }

tests/unit/calibration/test_compare_calibration_runs.py

@@ -15,25 +15,25 @@
 
 
 def test_run_comparison_paths_are_run_scoped():
-    paths = RunComparisonPaths("usdata-gha123-a1-abcdef12", version="1.73.0")
+    paths = RunComparisonPaths("usdata-gha123-a1", version="1.73.0")
 
     assert (
         paths.regional_diagnostics
         == "hf://policyengine/policyengine-us-data/calibration/runs/"
-        "usdata-gha123-a1-abcdef12/diagnostics/unified_diagnostics.csv"
+        "usdata-gha123-a1/diagnostics/unified_diagnostics.csv"
     )
     assert (
         paths.national_diagnostics
         == "hf://policyengine/policyengine-us-data/calibration/runs/"
-        "usdata-gha123-a1-abcdef12/diagnostics/national_unified_diagnostics.csv"
+        "usdata-gha123-a1/diagnostics/national_unified_diagnostics.csv"
     )
     assert (
         paths.candidate_h5 == "hf://policyengine/policyengine-us-data/staging/"
-        "1.73.0-usdata-gha123-a1-abcdef12/national/US.h5"
+        "1.73.0-usdata-gha123-a1/national/US.h5"
     )
     assert (
         paths.legacy_h5 == "hf://policyengine/policyengine-us-data/staging/"
-        "1.73.0-usdata-gha123-a1-abcdef12/enhanced_cps_2024.h5"
+        "1.73.0-usdata-gha123-a1/enhanced_cps_2024.h5"
     )
 
 

tests/unit/test_modal_local_area.py

@@ -51,13 +51,13 @@ def test_promote_scripts_can_defer_staging_cleanup_for_pipeline_promotion():
 
     regional_script = local_area._build_promote_publish_script(
         version="1.73.0",
-        run_id="usdata-gha123-a1-abcdef12",
+        run_id="usdata-gha123-a1",
         rel_paths=["states/AL.h5"],
         cleanup_staging=False,
     )
     national_script = local_area._build_promote_national_publish_script(
         version="1.73.0",
-        run_id="usdata-gha123-a1-abcdef12",
+        run_id="usdata-gha123-a1",
         rel_paths=["national/US.h5"],
         cleanup_staging=False,
     )
@@ -72,7 +72,7 @@ def test_promote_publish_falls_back_to_package_version_for_new_run_ids(
     monkeypatch, tmp_path
 ):
     local_area = load_local_area_module()
-    run_id = "usdata-gha123-a1-abcdef12"
+    run_id = "usdata-gha123-a1"
     run_dir = tmp_path / run_id
     run_dir.mkdir()
     (run_dir / "manifest.json").write_text(
@@ -107,7 +107,7 @@ def test_promote_national_publish_falls_back_to_package_version_for_new_run_ids(
     monkeypatch,
 ):
     local_area = load_local_area_module()
-    run_id = "usdata-gha123-a1-abcdef12"
+    run_id = "usdata-gha123-a1"
     captured = {}
 
     monkeypatch.setattr(local_area, "setup_gcp_credentials", lambda: None)

tests/unit/test_pipeline.py

@@ -13,6 +13,7 @@
     NATIONAL_FIT_LAMBDA_L0,
     _build_diagnostics_upload_script,
     _calibration_package_parameters,
+    _new_run_metadata,
     _pipeline_error_summary,
     _run_required_promotion_subprocess,
 )
@@ -21,6 +22,7 @@
     read_run_meta,
     write_run_meta,
 )
+from policyengine_us_data.utils.run_context import RunContext  # noqa: E402
 from policyengine_us_data.utils.step_manifest import ArtifactReference  # noqa: E402
 
 
@@ -102,6 +104,37 @@ def test_pipeline_error_summary_falls_back_to_bounded_traceback(monkeypatch):
     assert "old traceback" not in summary
 
 
+def test_new_run_metadata_accepts_release_context_fields_once():
+    context = RunContext.from_mapping(
+        {
+            "run_id": "run-123",
+            "candidate_version": "1.73.0-minor",
+            "release_version": "",
+            "base_release_version": "1.73.0",
+            "release_bump": "minor",
+            "modal_app_name": "us-data-1-73-0-minor-run-123",
+            "modal_environment": "main",
+            "hf_staging_prefix": "staging/1.73.0-minor-run-123",
+        }
+    )
+
+    meta = _new_run_metadata(
+        run_id=context.run_id,
+        branch="main",
+        sha="abc123",
+        candidate_version=context.candidate_version,
+        release_version=context.release_version,
+        run_context=context,
+    )
+
+    assert meta.base_release_version == "1.73.0"
+    assert meta.release_bump == "minor"
+    assert meta.modal_app_name == "us-data-1-73-0-minor-run-123"
+    assert meta.hf_staging_prefix == "staging/1.73.0-minor-run-123"
+    assert meta.run_context["base_release_version"] == "1.73.0"
+    assert meta.run_context["release_bump"] == "minor"
+
+
 class TestRunMetadata:
     def test_to_dict(self):
         meta = RunMetadata(

tests/unit/test_release_manifest.py

@@ -211,17 +211,17 @@ def test_build_release_manifest_records_run_context(tmp_path):
         version="1.73.0",
         repo_id="policyengine/policyengine-us-data",
         run_context={
-            "run_id": "usdata-gha123-a1-abcdef12",
-            "modal_app_name": "policyengine-us-data-pub-usdata-gha123-a1-abcdef12",
-            "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1-abcdef12",
+            "run_id": "usdata-gha123-a1",
+            "modal_app_name": "us-data-1-73-0-usdata-gha123-a1",
+            "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1",
         },
         created_at="2026-04-10T12:00:00Z",
     )
 
     assert manifest["build"]["metadata"]["run_context"] == {
-        "run_id": "usdata-gha123-a1-abcdef12",
-        "modal_app_name": "policyengine-us-data-pub-usdata-gha123-a1-abcdef12",
-        "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1-abcdef12",
+        "run_id": "usdata-gha123-a1",
+        "modal_app_name": "us-data-1-73-0-usdata-gha123-a1",
+        "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1",
     }
 
 
@@ -244,9 +244,9 @@ def test_build_release_manifest_validates_against_bundle_contract(tmp_path):
         version="1.73.0",
         repo_id="policyengine/policyengine-us-data",
         run_context={
-            "run_id": "usdata-gha123-a1-abcdef12",
-            "modal_app_name": "policyengine-us-data-pub-usdata-gha123-a1-abcdef12",
-            "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1-abcdef12",
+            "run_id": "usdata-gha123-a1",
+            "modal_app_name": "us-data-1-73-0-usdata-gha123-a1",
+            "hf_staging_prefix": "staging/1.73.0-usdata-gha123-a1",
         },
         model_package_version=EXPECTED_MODEL_PACKAGE_VERSION,
         model_package_git_sha="deadbeef",

tests/unit/test_remote_calibration_runner.py

@@ -200,7 +200,7 @@ def fake_run_streaming(cmd, env=None, label=""):
         workers=1,
         n_clones=10,
         run_id="bench-run",
-        modal_app_name="policyengine-us-data-pub-bench-run",
+        modal_app_name="us-data-bench-run",
         modal_environment="main",
         pipeline_volume_name="pipeline-artifacts-bench-run",
         chunked_matrix=True,
@@ -220,11 +220,8 @@ def fake_run_streaming(cmd, env=None, label=""):
     assert captured["cmd"][chunk_dir_idx] == str(artifacts_dir / "matrix_build")
     assert captured["env"]["POLICYENGINE_US_DATA_RUN_ID"] == "bench-run"
     assert captured["env"]["US_DATA_RUN_ID"] == "bench-run"
-    assert (
-        captured["env"]["US_DATA_MODAL_APP_NAME"]
-        == "policyengine-us-data-pub-bench-run"
-    )
-    assert captured["env"]["MODAL_APP_NAME"] == "policyengine-us-data-pub-bench-run"
+    assert captured["env"]["US_DATA_MODAL_APP_NAME"] == "us-data-bench-run"
+    assert captured["env"]["MODAL_APP_NAME"] == "us-data-bench-run"
     assert captured["env"]["US_DATA_MODAL_ENVIRONMENT"] == "main"
     assert captured["env"]["MODAL_ENVIRONMENT"] == "main"
     assert (