abhigyanpatwari/GitNexus · error · ValueError

unsafe Git reflog metadata blocks oracle sanitization

Error message

unsafe Git reflog metadata blocks oracle sanitization

What it means

Before `shutil.rmtree(.git/logs)`, the harness verifies .git/logs is a real directory and not a symlink. Calling rmtree on a symlinked logs would follow the link and delete arbitrary tree outside .git, so a symlink (or non-directory) logs is treated as unsafe.

Source

Thrown at eval/workflow_bench/oracle_assets.py:408

        "ORIG_HEAD",
        "REBASE_HEAD",
        "REVERT_HEAD",
        "shallow",
    ):
        path = git_dir / pseudo_ref
        try:
            metadata = path.lstat()
        except FileNotFoundError:
            continue
        if stat.S_ISLNK(metadata.st_mode) or not stat.S_ISREG(metadata.st_mode):
            raise ValueError(f"unsafe Git metadata blocks oracle sanitization: {pseudo_ref}")
        path.unlink()

    logs = git_dir / "logs"
    if logs.exists() or logs.is_symlink():
        logs_metadata = logs.lstat()
        if stat.S_ISLNK(logs_metadata.st_mode) or not stat.S_ISDIR(logs_metadata.st_mode):
            raise ValueError("unsafe Git reflog metadata blocks oracle sanitization")
        shutil.rmtree(logs)

    _git_checked(root, ["repack", "-A", "-d"], timeout=600)
    _git_checked(root, ["prune", "--expire=now"], timeout=600)
    _git_checked(root, ["prune-packed"], timeout=600)

    remaining_refs = _git_checked(root, ["for-each-ref", "--format=%(refname)"], timeout=60)
    if remaining_refs:
        raise ValueError("oracle sanitization left clone references recoverable")
    fsck = run_checked(
        ["git", "-C", str(root), "fsck", "--full", "--no-progress", "--no-reflogs", "--unreachable"],
        timeout=600,
        tail_bytes=MAX_CLONE_REF_BYTES,
    )
    if fsck.stdout_tail.strip() or fsck.stderr_tail.strip():
        raise ValueError("oracle sanitization left unreachable Git objects recoverable")

    forbidden_objects: list[tuple[str, str]] = []

View on GitHub (pinned to d540b00184)

Solutions

1. Inspect `ls -la <clone>/.git/logs`; if it is a symlink, remove the link: `rm <clone>/.git/logs`.
2. If it is a regular file, delete it: `rm <clone>/.git/logs` (git will recreate a real directory when needed).
3. Re-clone from a trusted source if the layout looks crafted.

import stat
from pathlib import Path

def logs_is_real_dir_or_absent(clone: Path) -> bool:
    logs = clone / ".git" / "logs"
    try:
        st = logs.lstat()
    except FileNotFoundError:
        return True
    return not stat.S_ISLNK(st.st_mode) and stat.S_ISDIR(st.st_mode)

def is_unsafe_reflog_metadata(exc: BaseException) -> bool:
    return isinstance(exc, ValueError) and "unsafe Git reflog metadata" in str(exc)

try:
    oracle_assets.sanitize_clone_for_hidden_oracles(clone)
except ValueError as exc:
    quarantine(clone)
    raise AbortTask(str(exc)) from exc

When it happens

Trigger: Triggered when .git/logs exists and lstat reports it as a symlink or non-directory (e.g., a regular file, or a symlink pointing outside the repo).

Common situations: A crafted clone that symlinks .git/logs to /tmp or another repo; a clone restored from a tarball that did not preserve directory-ness of .git/logs; a filesystem where logs was replaced by a file.

Related errors

• unsafe Git metadata blocks oracle sanitization: {pseudo_ref}
• oracle sanitization requires a real self-contained clone: {r
• benchmark harness checkout must contain only real directorie
• clone contains an unsafe reference name
• clone contains unsafe or unbounded remote metadata

AI-assisted analysis of abhigyanpatwari/GitNexus@d540b00184 (2026-08-12). Data as JSON: /api/errors/22c32f6f98ddcd92. Report an issue: GitHub.