fix(skills): stop syncing bookkeeping dirs to sandboxes

iter_skills_files() walked the skills tree with a bare rglob("*"), so the
.hub download cache, .archive, curator backups, and any node_modules/.git
under a skill package were uploaded to the sandbox on every sync. The
sandbox never reads them: skill content is resolved host-side.

EXCLUDED_SKILL_DIRS is already the canonical exclusion set, honoured by
discovery and backup. Apply it to the sync path too, across all three
roots iter_skills_files() walks (local, external, project-local), and add
.curator_backups to the set.

Measured on a local install: 900 files / 67.3 MB -> 771 files / 8.4 MB.

This is not just wasted bandwidth on the SSH backend, where the oversized
payload can exceed the 120s _ssh_bulk_upload deadline and surface as the
agent hanging on every tool call.

The filter intentionally does not reuse is_excluded_skill_path(), which
also prunes references/, templates/, assets/ and scripts/ -- those hold
support files and bundled scripts the sandbox does read and execute.

agent/skill_utils.py

@@ -31,6 +31,7 @@ EXCLUDED_SKILL_DIRS = frozenset(
         ".github",
         ".hub",
         ".archive",
+        ".curator_backups",
         ".venv",
         "venv",
         "node_modules",

tests/tools/test_credential_files.py

@@ -147,6 +147,46 @@ class TestIterSkillsFiles:
         # Symlink should be excluded
         assert not any("evil" in f["container_path"] for f in files)
 
+    def test_skips_excluded_bookkeeping_dirs(self, tmp_path):
+        """Bookkeeping and dependency dirs must not be uploaded to a sandbox.
+
+        The sync path used a bare rglob("*"), so the .hub download cache,
+        .archive, curator backups and any node_modules/.git under a skills
+        tree were packed up on every sync even though the sandbox never
+        reads them. Sync now honours EXCLUDED_SKILL_DIRS like discovery.
+        """
+        hermes_home = tmp_path / ".hermes"
+        skills_dir = hermes_home / "skills"
+        (skills_dir / "cat" / "myskill").mkdir(parents=True)
+        (skills_dir / "cat" / "myskill" / "SKILL.md").write_text("# skill")
+        # Progressive-disclosure support files must still be synced.
+        (skills_dir / "cat" / "myskill" / "references").mkdir()
+        (skills_dir / "cat" / "myskill" / "references" / "api.md").write_text("ref")
+
+        for excluded in (".hub", ".archive", ".curator_backups", "node_modules"):
+            junk = skills_dir / excluded / "vendored"
+            junk.mkdir(parents=True)
+            (junk / "SKILL.md").write_text("# stale copy")
+        # Also nested inside an otherwise-valid skill package.
+        cache = skills_dir / "cat" / "myskill" / "__pycache__"
+        cache.mkdir()
+        (cache / "helper.cpython-311.pyc").write_text("bytecode")
+
+        with patch.dict(os.environ, {"HERMES_HOME": str(hermes_home)}):
+            files = iter_skills_files()
+
+        paths = {f["container_path"] for f in files}
+        assert "/root/.hermes/skills/cat/myskill/SKILL.md" in paths
+        assert "/root/.hermes/skills/cat/myskill/references/api.md" in paths
+        for excluded in (
+            ".hub",
+            ".archive",
+            ".curator_backups",
+            "node_modules",
+            "__pycache__",
+        ):
+            assert not any(excluded in path for path in paths), excluded
+
     def test_empty_when_no_skills_dir(self, tmp_path):
         hermes_home = tmp_path / ".hermes"
         hermes_home.mkdir()

tools/credential_files.py

@@ -28,6 +28,8 @@ from pathlib import Path
 from typing import Dict, List, Optional
 from hermes_cli.config import cfg_get
 
+from agent.skill_utils import EXCLUDED_SKILL_DIRS
+
 try:  # pragma: no cover - exercised via the fail-closed test below
     from agent.file_safety import get_read_block_error
 except ImportError:  # noqa: F401 - sentinel consumed in register_credential_file
@@ -344,15 +346,32 @@ def _safe_skills_path(skills_dir: Path) -> str:
     return str(safe_dir)
 
 
+def _is_excluded_skills_dir(rel: Path) -> bool:
+    """True if *rel* sits under a directory excluded from skill scanning.
+
+    Reuses ``agent.skill_utils.EXCLUDED_SKILL_DIRS`` so the sync path agrees
+    with discovery. These are local bookkeeping and dependency directories
+    (``.hub`` download cache, ``.archive``, ``.curator_backups``,
+    ``node_modules``, ``__pycache__``, ``.git``, ...) that the remote agent
+    never reads.
+
+    This deliberately does not use ``is_excluded_skill_path()``, which also
+    prunes ``references/``, ``templates/``, ``assets/`` and ``scripts/``.
+    Those hold progressive-disclosure support files and bundled scripts the
+    sandbox does execute, so they must keep syncing.
+    """
+    return any(part in EXCLUDED_SKILL_DIRS for part in rel.parts[:-1])
+
+
 def iter_skills_files(
     container_base: str = "/root/.hermes",
 ) -> List[Dict[str, str]]:
     """Yield individual (host_path, container_path) entries for skills files.
 
     Includes both the local skills dir and any external dirs configured via
-    skills.external_dirs.  Skips symlinks entirely.  Preferred for backends
-    that upload files individually (Daytona, Modal) rather than mounting a
-    directory.
+    skills.external_dirs.  Skips symlinks and anything under
+    EXCLUDED_SKILL_DIRS entirely.  Preferred for backends that upload files
+    individually (Daytona, Modal) rather than mounting a directory.
     """
     result: List[Dict[str, str]] = []
 
@@ -364,6 +383,8 @@ def iter_skills_files(
             if item.is_symlink() or not item.is_file():
                 continue
             rel = item.relative_to(skills_dir)
+            if _is_excluded_skills_dir(rel):
+                continue
             result.append({
                 "host_path": str(item),
                 "container_path": f"{container_root}/{rel}",
@@ -380,6 +401,8 @@ def iter_skills_files(
                 if item.is_symlink() or not item.is_file():
                     continue
                 rel = item.relative_to(ext_dir)
+                if _is_excluded_skills_dir(rel):
+                    continue
                 result.append({
                     "host_path": str(item),
                     "container_path": f"{container_root}/{rel}",
@@ -392,6 +415,8 @@ def iter_skills_files(
                 if item.is_symlink() or not item.is_file():
                     continue
                 rel = item.relative_to(proj_dir)
+                if _is_excluded_skills_dir(rel):
+                    continue
                 result.append({
                     "host_path": str(item),
                     "container_path": f"{container_root}/{rel}",