merge: resolve PR #481 conflicts with main

Author: jdchawla29

docs/v6/reference/runtime.mdx

@@ -153,10 +153,12 @@ DaytonaRuntime(snapshot_name=None, *, image=None, command=None, workdir="/app",
 ```
 
 - **`snapshot_name`** - Daytona snapshot to boot from (the durable handle).
-- **`image`** - Dockerfile/registry ref to build the snapshot once if it's missing. Resources (cpu/memory/gpu) live on the snapshot.
+- **`image`** - Dockerfile/registry ref to build the snapshot if it's missing. Daytona records what a snapshot was built from, so when the image content changes (an edited Dockerfile or context file, a repointed registry ref) the snapshot is rebuilt in place under the same name instead of silently reusing the build from before the edit.
 - **`workdir`** / **`port`** - guest working directory and in-sandbox serving port.
 - **`ssh_host`** / **`ssh_expires_minutes`** - SSH tunnel settings (Daytona exposes services over an SSH local-forward).
 
+Resources (cpu/memory/gpu) are fixed on the snapshot at build time. With `image`, a task's `runtime_config.resources` builds a sized variant under a suffixed name (`my-env-4cpu`); without `image`, an already-built snapshot cannot be resized.
+
 ### `HUDRuntime`
 
 ```python

hud/agents/tests/test_provider_native_tools.py

@@ -10,6 +10,7 @@
 import shlex
 from typing import Any, cast
 
+import asyncssh
 import mcp.types as mcp_types
 import pytest
 
@@ -47,7 +48,22 @@ async def run(
         self.commands.append(command)
         parts = shlex.split(command)
         if len(parts) == 3 and parts[:2] == ["cat", "--"]:
-            return _Completed(stdout=self._store.get(parts[2], b"").decode())
+            if parts[2] not in self._store:
+                if check:
+                    raise asyncssh.ProcessError(
+                        env=None,
+                        command=command,
+                        subsystem=None,
+                        exit_status=1,
+                        exit_signal=None,
+                        returncode=1,
+                        stdout="",
+                        stderr=f"cat: {parts[2]}: No such file or directory",
+                    )
+                return _Completed(
+                    stderr=f"cat: {parts[2]}: No such file or directory", exit_status=1
+                )
+            return _Completed(stdout=self._store[parts[2]].decode())
         if len(parts) == 3 and parts[:2] == ["cat", ">"]:
             assert input is not None
             self._store[parts[2]] = input.encode()
@@ -520,3 +536,37 @@ async def test_gemini_edit_creates_file_when_old_string_empty() -> None:
     await tool.execute({"file_path": "/n.txt", "old_string": "", "new_string": "fresh"})
 
     assert ssh.files["/n.txt"] == b"fresh"
+
+
+def test_map_path_leaves_a_symlinked_spelling_of_the_workspace_alone() -> None:
+    """A workspace made at /tmp/w is served as /private/tmp/w on macOS; re-anchoring
+    the caller's spelling instead of stripping it nests the path under itself."""
+    cap = Capability(
+        name="shell",
+        protocol="ssh/2",
+        url="ssh://localhost:22",
+        params={"cwd": "/private/tmp/w", "cwd_aliases": ["/tmp/w"]},
+    )
+    ssh = SSHClient(cap, cast("Any", None))
+
+    assert ssh.map_path("/tmp/w/calc.py") == "/private/tmp/w/calc.py"
+    assert ssh.map_path("/private/tmp/w/calc.py") == "/private/tmp/w/calc.py"
+    assert ssh.map_path("/tmp/w") == "/private/tmp/w"
+    # Workspace-relative addressing still anchors, and an unrelated absolute
+    # path is still clamped into the workspace like a chroot.
+    assert ssh.map_path("/REPORT.md") == "/private/tmp/w/REPORT.md"
+    assert ssh.map_path("/tmp/elsewhere/f.txt") == "/private/tmp/w/tmp/elsewhere/f.txt"
+
+
+async def test_reading_a_missing_file_is_a_tool_error_not_a_raised_traceback() -> None:
+    """Reading before creating is the first thing an editor tool does; that failure
+    must come back as a tool result carrying the shell's message."""
+    ssh = _FakeSSH()
+    tool = ClaudeTextEditorTool(
+        spec=ClaudeTextEditorTool.default_spec("claude"), client=cast("SSHClient", ssh)
+    )
+
+    result = await tool.execute({"command": "view", "path": "/nope.txt"})
+
+    assert result.isError is True
+    assert "No such file or directory" in result_text(result)

hud/agents/tools/ssh.py

@@ -7,13 +7,22 @@
 
 from __future__ import annotations
 
+import asyncssh
 import mcp.types as mcp_types
 
-from hud.agents.tools.base import AgentTool
+from hud.agents.tools.base import AgentTool, tool_err, tool_ok
 from hud.capabilities import SSHClient
 from hud.types import MCPToolResult
 
 
+def _remote_error(exc: asyncssh.ProcessError) -> str:
+    """What the remote command printed to stderr — a failed file op is an
+    ordinary tool outcome, so the agent gets the shell's message, not the
+    exception's repr."""
+    stderr = exc.stderr.decode("utf-8", "replace") if isinstance(exc.stderr, bytes) else exc.stderr
+    return (stderr or "").strip() or f"exit {exc.exit_status}"
+
+
 class SSHTool(AgentTool[SSHClient]):
     """Capability base: tool driven by an ``SSHClient``."""
 
@@ -37,19 +46,26 @@ async def bash(self, command: str) -> MCPToolResult:
 
     async def file_read(self, path: str) -> MCPToolResult:
         """Read a text file through SSH exec."""
-        return tool_ok(await self.client.read_text(path))
+        try:
+            return tool_ok(await self.client.read_text(path))
+        except asyncssh.ProcessError as e:
+            return tool_err(_remote_error(e))
 
     async def file_write(self, path: str, content: str) -> MCPToolResult:
         """Write a text file through SSH exec."""
-        await self.client.write_text(path, content)
+        try:
+            await self.client.write_text(path, content)
+        except asyncssh.ProcessError as e:
+            return tool_err(_remote_error(e))
         return tool_ok(f"wrote {len(content)} bytes to {path}")
 
     async def file_list(self, path: str = "/") -> MCPToolResult:
         """List directory entries through SSH exec."""
-        names = await self.client.listdir(path)
+        try:
+            names = await self.client.listdir(path)
+        except asyncssh.ProcessError as e:
+            return tool_err(_remote_error(e))
         return tool_ok("\n".join(names) if names else "(empty)")
 
 
-from hud.agents.tools.base import tool_ok  # noqa: E402
-
 __all__ = ["SSHTool"]

hud/capabilities/base.py

@@ -7,9 +7,12 @@
 import sys
 from abc import ABC, abstractmethod
 from dataclasses import dataclass, field
-from typing import Any, ClassVar, Self
+from typing import TYPE_CHECKING, Any, ClassVar, Self
 from urllib.parse import urlsplit
 
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
 #: Matches the scheme prefix of a URL (RFC 3986).
 SCHEME_RE: re.Pattern[str] = re.compile(r"^([a-zA-Z][a-zA-Z0-9+\-.]*):")
 
@@ -79,6 +82,7 @@ def ssh(
         client_key_path: str | os.PathLike[str] | None = None,
         shell: str | None = None,
         cwd: str | None = None,
+        cwd_aliases: Sequence[str] | None = None,
     ) -> Capability:
         """``ssh/2`` — SSH daemon with publickey auth.
 
@@ -90,6 +94,9 @@ def ssh(
         from ``sys.platform`` at construction time. Agents read this to
         format commands correctly. ``cwd`` is the absolute path sessions
         start in (the served workspace); clients anchor file paths to it.
+        ``cwd_aliases`` are other names the same directory answers to (paths
+        that reach it through symlinks), which clients treat as already
+        anchored rather than as workspace-relative addresses.
         """
         normalized = normalize_url(url, default_scheme="ssh", default_port=22)
         if shell is None:
@@ -101,6 +108,8 @@ def ssh(
             params["client_key_path"] = os.fspath(client_key_path)
         if cwd is not None:
             params["cwd"] = cwd
+        if cwd_aliases:
+            params["cwd_aliases"] = list(cwd_aliases)
         return cls(name=name, protocol="ssh/2", url=normalized, params=params)
 
     @classmethod

hud/capabilities/ssh.py

@@ -56,10 +56,17 @@ def map_path(self, path: str) -> str:
         real filesystem; replicate the chroot: strip the cwd prefix if present,
         normalize the remainder against ``/`` (clamping ``..`` at the root,
         exactly as a chroot does), and re-anchor under the cwd. Idempotent.
+
+        ``cwd_aliases`` are the same directory reached through symlinks, and are
+        stripped like the cwd: re-anchoring one would turn an already correct
+        absolute path into a nested one that does not exist.
         """
         cwd = str(self.capability.params.get("cwd", "")).rstrip("/")
         if not cwd:
             return path
+        aliases = [
+            str(alias).rstrip("/") for alias in self.capability.params.get("cwd_aliases") or []
+        ]
         if self._is_windows:
             # The workspace publishes cwd via as_posix() (e.g. "C:/work") but
             # callers pass native paths ("C:\work\file.txt"); NTFS paths are
@@ -70,8 +77,11 @@ def map_path(self, path: str) -> str:
             elif len(path) >= 2 and path[1] == ":" and path[0].isalpha():
                 # Drive-absolute outside the workspace: anchor like the chroot.
                 path = path[2:]
-        elif path == cwd or path.startswith(cwd + "/"):
-            path = path[len(cwd) :]
+        else:
+            for prefix in (cwd, *aliases):
+                if prefix and (path == prefix or path.startswith(prefix + "/")):
+                    path = path[len(prefix) :]
+                    break
         normalized = posixpath.normpath("/" + path.lstrip("/"))
         return cwd if normalized == "/" else cwd + normalized
 

hud/cli/trace.py

@@ -161,6 +161,8 @@ def _load_remote(trace_id: str) -> list[dict[str, Any]]:
 
 
 def _render_events(events: list[dict[str, Any]]) -> None:
+    # Payloads render as Text, never as markup: rich would read a literal
+    # `[len(s) // 2]` in agent output as a style tag and drop it.
     turn = 0
     for ev in events:
         kind = ev.get("kind")
@@ -175,7 +177,7 @@ def _render_events(events: list[dict[str, Any]]) -> None:
 
             text = ev.get("text")
             if text:
-                console.print(text)
+                console.print(Text(str(text)))
 
             for tc in ev.get("tool_calls") or []:
                 name = tc.get("name") or tc.get("function", {}).get("name", "?")
@@ -184,29 +186,30 @@ def _render_events(events: list[dict[str, Any]]) -> None:
                     with contextlib.suppress(Exception):
                         args = json.loads(args)
                 console.print(
-                    f"  [green]→[/green] [bold]{name}[/bold]({_fmt_args(args)})",
-                    highlight=False,
+                    Text.assemble(
+                        "  ", ("→", "green"), " ", (str(name), "bold"), f"({_fmt_args(args)})"
+                    )
                 )
 
             if ev.get("error"):
-                console.print(f"  [red]error: {ev['error']}[/red]")
+                console.print(Text(f"  error: {ev['error']}", style="red"))
 
         elif kind in ("tool_call", "tool_result"):
             name = ev.get("tool_name") or ev.get("name") or "?"
             result = ev.get("result_text") or ev.get("result") or ""
             error = ev.get("error")
             if error:
-                console.print(f"  [red]✗ {name}: {error}[/red]")
+                console.print(Text(f"  ✗ {name}: {error}", style="red"))
             else:
-                console.print(f"  [dim]{name} →[/dim]")
+                console.print(Text(f"  {name} →", style="dim"))
                 for line in str(result).splitlines():
-                    console.print(f"    {line}")
+                    console.print(Text(f"    {line}"))
 
         elif kind == "environment":
             msg = ev.get("text") or ev.get("content") or ""
             if msg:
                 console.print(Rule("[yellow]env[/yellow]", style="yellow"))
-                console.print(msg)
+                console.print(Text(str(msg)))
 
 
 def _fmt_args(args: Any) -> str:

hud/environment/tests/test_workspace.py

@@ -6,6 +6,7 @@
 import sys
 import tempfile
 from pathlib import Path
+from typing import Any, cast
 
 import asyncssh
 import pytest
@@ -268,3 +269,35 @@ def test_required_isolation_refuses_when_unavailable(monkeypatch, tmp_path) -> N
 
     with pytest.raises(RuntimeError, match="isolation was required"):
         ws.Workspace(tmp_path, require_isolation=True)
+
+
+@pytest.mark.asyncio
+async def test_a_symlinked_root_publishes_both_spellings(tmp_path: Path) -> None:
+    """A workspace addressed through a symlink (macOS /tmp -> /private/tmp) serves the
+    real path, so it must publish the caller's spelling too or clients re-anchor it."""
+    real = tmp_path / "real"
+    real.mkdir()
+    link = tmp_path / "link"
+    link.symlink_to(real, target_is_directory=True)
+
+    ws = Workspace(link)
+    await ws.start()
+    try:
+        cap = ws.capability()
+        assert cap.params["cwd"] == real.as_posix()
+        assert cap.params["cwd_aliases"] == [link.as_posix()]
+
+        client = SSHClient(cap, cast("Any", None))
+        assert client.map_path(f"{link}/calc.py") == f"{real}/calc.py"
+    finally:
+        await ws.stop()
+
+
+@pytest.mark.asyncio
+async def test_a_plain_root_publishes_no_alias(tmp_path: Path) -> None:
+    ws = Workspace(tmp_path / "root")
+    await ws.start()
+    try:
+        assert "cwd_aliases" not in ws.capability().params
+    finally:
+        await ws.stop()

hud/environment/workspace.py

@@ -203,6 +203,13 @@ def __init__(
         # Only override the default; respect an explicit guest_path.
         if self._bwrap is None and guest_path == "/workspace":
             self._guest_path = self.root.as_posix()
+        # The caller's spelling of the same directory when it differs from the real
+        # path (macOS resolves /tmp to /private/tmp) and sessions run under the real
+        # one. A client that knows only the real path treats the other spelling as a
+        # workspace-relative address and re-anchors it somewhere that does not exist.
+        given = Path(root).absolute().as_posix()
+        real = self.root.as_posix()
+        self._cwd_aliases = [given] if given != real and self._guest_path == real else []
         # ssh config
         self._ssh_host = host
         self._ssh_port = port
@@ -430,6 +437,7 @@ def capability(self, name: str = "shell") -> Capability:
             client_key=key_path.read_text() if key_path else None,
             client_key_path=key_path,
             cwd=self._guest_path,
+            cwd_aliases=self._cwd_aliases or None,
         )
 
     @property

hud/eval/job.py

@@ -105,18 +105,26 @@ async def trace_enter(
     *,
     job_id: str | None,
     group_id: str | None,
+    task_slug: str,
     model: str | None = None,
 ) -> None:
     """Report that one rollout started.
 
-    ``model`` is the model string the agent will sample (when known); the
-    platform resolves it and attributes the trace immediately on enter.
+    ``task_slug`` identifies the logical task independently of the execution
+    environment. ``model`` is the model string the agent will sample (when
+    known); the platform resolves it and attributes the trace immediately on
+    enter.
     """
     if not _reporting_enabled():
         return
     await _report(
         f"/trace/{trace_id}/enter",
-        {"job_id": job_id, "group_id": group_id, "model": model},
+        {
+            "job_id": job_id,
+            "group_id": group_id,
+            "task_slug": task_slug,
+            "model": model,
+        },
     )