feat(registration): report the agent's commit and source repo at registration

src/agentex/lib/core/tracing/code_revision.py

@@ -21,7 +21,7 @@
 
 from agentex.lib.utils.logging import make_logger
 
-__all__ = ("COMMIT_SHA_KEY", "enable", "disable", "is_enabled", "commit_sha")
+__all__ = ("COMMIT_SHA_KEY", "enable", "disable", "is_enabled", "commit_sha", "is_git_object_name")
 
 logger = make_logger(__name__)
 
@@ -31,6 +31,12 @@
 # git's own 7-character minimum.
 _GIT_SHA_RE = re.compile(r"[0-9a-fA-F]{7,64}")
 
+
+def is_git_object_name(value: str) -> bool:
+    """Whether ``value`` is a full or abbreviated git SHA-1/SHA-256 object name."""
+    return _GIT_SHA_RE.fullmatch(value.strip()) is not None
+
+
 _COMMIT_SHA_ENV = "AGENT_COMMIT_SHA"
 # Fallback only: automatic, and only usable when it happens to be SHA-shaped.
 _AGENT_VERSION_ENV = "AGENT_VERSION"

src/agentex/lib/environment_variables.py

@@ -26,6 +26,7 @@ class EnvVarKeys(str, Enum):
     AGENT_ID = "AGENT_ID"
     AGENT_VERSION = "AGENT_VERSION"
     AGENT_COMMIT_SHA = "AGENT_COMMIT_SHA"
+    AGENT_SOURCE_REPO = "AGENT_SOURCE_REPO"
     AGENT_API_KEY = "AGENT_API_KEY"
     # ACP Configuration
     ACP_URL = "ACP_URL"
@@ -74,6 +75,8 @@ class EnvironmentVariables(BaseModel):
     # `adk.code_revision.enable()`, which also refuses a value that is not a git
     # object name. See agentex.lib.core.tracing.code_revision.
     AGENT_COMMIT_SHA: str | None = None
+    # Git remote the agent was built from (any URL form; normalized to host/path on use).
+    AGENT_SOURCE_REPO: str | None = None
     AGENT_API_KEY: str | None = None
     ACP_TYPE: str | None = "async"
     AGENT_INPUT_TYPE: str | None = None

src/agentex/lib/utils/build_provenance.py

@@ -82,7 +82,8 @@ def normalize_remote(url: Optional[str]) -> Optional[str]:
     """Strip credentials and scheme from a remote, returning ``host/path``."""
     if not url:
         return None
-    candidate = url.strip()
+    # Query strings and fragments never name a repo, but they do carry tokens.
+    candidate = url.strip().split("?", 1)[0].split("#", 1)[0]
     # scp-like syntax: git@host:org/repo(.git) — no scheme, host/path split on ':'
     if "://" not in candidate and ":" in candidate and "/" not in candidate.split(":", 1)[0]:
         candidate = candidate.split("@", 1)[-1].replace(":", "/", 1)

src/agentex/lib/utils/registration.py

@@ -7,6 +7,8 @@
 
 from agentex.lib.utils.logging import make_logger
 from agentex.lib.environment_variables import EnvironmentVariables
+from agentex.lib.utils.build_provenance import normalize_remote
+from agentex.lib.core.tracing.code_revision import is_git_object_name
 
 logger = make_logger(__name__)
 
@@ -20,6 +22,29 @@ def get_auth_principal(env_vars: EnvironmentVariables):
     except Exception:
         return None
 
+
+def build_registration_metadata(env_vars: EnvironmentVariables, agent_card=None) -> dict:
+    """Deployment id, source provenance, and agent card; keys appear only when known."""
+    metadata: dict = {}
+    if env_vars.AGENTEX_DEPLOYMENT_ID:
+        metadata["deployment_id"] = env_vars.AGENTEX_DEPLOYMENT_ID
+    commit = (env_vars.AGENT_COMMIT_SHA or "").strip()
+    if commit:
+        if is_git_object_name(commit):
+            metadata["commit_sha"] = commit
+        else:
+            logger.warning(
+                "AGENT_COMMIT_SHA=%r is not a git commit SHA; commit_sha omitted from registration.",
+                commit,
+            )
+    repo = normalize_remote(env_vars.AGENT_SOURCE_REPO)
+    if repo:
+        metadata["source_repo"] = repo
+    if agent_card is not None:
+        metadata["agent_card"] = agent_card.model_dump() if hasattr(agent_card, "model_dump") else agent_card
+    return metadata
+
+
 async def register_agent(env_vars: EnvironmentVariables, agent_card=None):
     """Register this agent with the Agentex server"""
     if not env_vars.AGENTEX_BASE_URL:
@@ -33,13 +58,7 @@ async def register_agent(env_vars: EnvironmentVariables, agent_card=None):
         or f"Generic description for agent: {env_vars.AGENT_NAME}"
     )
 
-    # Registration metadata carries the deployment id and agent card.
-    registration_metadata: dict = {}
-    if env_vars.AGENTEX_DEPLOYMENT_ID:
-        registration_metadata["deployment_id"] = env_vars.AGENTEX_DEPLOYMENT_ID
-    if agent_card is not None:
-        card_data = agent_card.model_dump() if hasattr(agent_card, "model_dump") else agent_card
-        registration_metadata["agent_card"] = card_data
+    registration_metadata = build_registration_metadata(env_vars, agent_card)
 
     # Prepare registration data
     registration_data = {

tests/lib/test_agent_card.py

@@ -333,6 +333,8 @@ def mock_env_vars(self):
             "AGENT_ID": None,
             "AGENT_INPUT_TYPE": None,
             "AGENT_API_KEY": None,
+            "AGENT_COMMIT_SHA": None,
+            "AGENT_SOURCE_REPO": None,
             "AGENTEX_DEPLOYMENT_ID": None,
         })()
         return mock

tests/lib/test_build_provenance.py

@@ -50,6 +50,7 @@ def _write(root: Path, rel: str, content: str = "x") -> None:
         ("https://github.com/scaleapi/Repo.git", "github.com/scaleapi/Repo"),
         ("https://x-token:secret@GitHub.com/scaleapi/Repo", "github.com/scaleapi/Repo"),
         ("ssh://git@gitlab.com/group/sub/proj.git", "gitlab.com/group/sub/proj"),
+        ("https://github.com/scaleapi/Repo.git?access_token=SECRET#frag", "github.com/scaleapi/Repo"),
         ("", None),
         (None, None),
     ],

tests/lib/utils/test_registration.py

@@ -0,0 +1,49 @@
+"""Registration metadata: what an agent reports about itself at startup."""
+
+from __future__ import annotations
+
+import pytest
+
+from agentex.lib.utils.registration import build_registration_metadata
+from agentex.lib.environment_variables import EnvironmentVariables
+
+SHA = "b362b171a9c4e1f09d8e7a6b5c4d3e2f1a0b9c8d"
+
+
+def _env(**overrides) -> EnvironmentVariables:
+    return EnvironmentVariables(AGENT_NAME="sample-agent", ACP_URL="http://agent", **overrides)
+
+
+def test_nothing_known_yields_empty_metadata():
+    assert build_registration_metadata(_env()) == {}
+
+
+def test_commit_and_repo_reported_when_set():
+    env = _env(AGENT_COMMIT_SHA=SHA, AGENT_SOURCE_REPO="git@github.com:scaleapi/Demo.git")
+    assert build_registration_metadata(env) == {
+        "commit_sha": SHA,
+        "source_repo": "github.com/scaleapi/Demo",
+    }
+
+
+@pytest.mark.parametrize("value", ["latest", "v1.2.3", "rocket_mock_agent-" + SHA, "abc", "   "])
+def test_non_commit_values_are_omitted_not_forwarded(value):
+    """A field named for a commit never holds an image tag, same rule as __commit_sha__."""
+    assert "commit_sha" not in build_registration_metadata(_env(AGENT_COMMIT_SHA=value))
+
+
+def test_repo_normalization_strips_scheme_and_credentials():
+    env = _env(AGENT_SOURCE_REPO="https://x-token:secret@GitHub.com/scaleapi/Demo.git")
+    assert build_registration_metadata(env)["source_repo"] == "github.com/scaleapi/Demo"
+
+
+def test_deployment_id_and_agent_card_still_reported():
+    class Card:
+        def model_dump(self):
+            return {"name": "sample"}
+
+    env = _env(AGENTEX_DEPLOYMENT_ID="dep-1")
+    assert build_registration_metadata(env, Card()) == {
+        "deployment_id": "dep-1",
+        "agent_card": {"name": "sample"},
+    }