dejima/dejima-ci-hardener
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Name the tiers after what a job gets, and put the reader first

Browse Source 
The README grew alongside the tool and ended up in the order the work
happened rather than the order a reader needs. Three things were wrong with
it. The two-tier split was implicit, so a reader had to sort out which
paragraphs applied to the host they were building. The opening defined the
whole tool as "no Docker socket in job containers, and deny-by-default
egress", which describes one tier only - someone building the other was told
it does things it deliberately will not do. And long-form justification sat
between the reader and the steps.

So the README is now tier-first: pick a tier from a comparison table, then
follow only that tier's half. Everything that is background rather than
instruction moved to docs/reference.md, with docs/README.md indexing it and
the Proxmox runbook. The runbook stays in proxmox/ beside the firewall
template it explains.

The tier names changed with it. `ci` and `runtime` were dejima's words for
these hosts: `ci` distinguishes nothing on a tool whose whole subject is CI,
and `runtime` only means something if you already know dejima's runtime tests
are the workload that forced the tier to exist. They are `confined` and
`privileged` now, after what a job on each is granted. The distinction that
generalises is where the security boundary sits - inside the host for one,
on the hypervisor for the other - and the table leads with it.

`unprivileged`/`privileged` would have been the more standard pairing and was
rejected: two characters apart on a flag where a misread silently produces a
host that looks hardened and is not.

The old --role values still work and say what they are now called rather than
translating silently. examples/ is renamed to match on both sides, so no
file's tier is implied by the absence of a suffix, and the config mounts in
the compose files follow.

Written down explicitly for the first time: the privileged tier has no egress
allowlist by design. Its jobs hold the socket and run in the host's network
namespace, so any filter installed there is one they can flush. Installing it
anyway would leave a host that looks like it has an allowlist, which is worse
than one that plainly does not.

Also adds GPL-3.0-or-later to match dejima: LICENSE plus SPDX headers on the
scripts.

bats test/ passes, 21/21. All four --role values and an invalid one behave as
intended; the compose and act_runner examples still parse.
This commit is contained in:
Brad Anderson
2026-08-06 19:31:55 -07:00
parent 45fad53f2a
commit f6156c2888
15 changed files with 1132 additions and 318 deletions
Download Patch File Download Diff File Expand all files Collapse all files
LICENSE
+674
View File
@@ -0,0 +1,674 @@
GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
The GNU General Public License is a free, copyleft license for
software and other kinds of works.
The licenses for most software and other practical works are designed
to take away your freedom to share and change the works. By contrast,
the GNU General Public License is intended to guarantee your freedom to
share and change all versions of a program--to make sure it remains free
software for all its users. We, the Free Software Foundation, use the
GNU General Public License for most of our software; it applies also to
any other work released this way by its authors. You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
them if you wish), that you receive source code or can get it if you
want it, that you can change the software or use pieces of it in new
free programs, and that you know you can do these things.
To protect your rights, we need to prevent others from denying you
these rights or asking you to surrender the rights. Therefore, you have
certain responsibilities if you distribute copies of the software, or if
you modify it: responsibilities to respect the freedom of others.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must pass on to the recipients the same
freedoms that you received. You must make sure that they, too, receive
or can get the source code. And you must show them these terms so they
know their rights.
Developers that use the GNU GPL protect your rights with two steps:
(1) assert copyright on the software, and (2) offer you this License
giving you legal permission to copy, distribute and/or modify it.
For the developers' and authors' protection, the GPL clearly explains
that there is no warranty for this free software. For both users' and
authors' sake, the GPL requires that modified versions be marked as
changed, so that their problems will not be attributed erroneously to
authors of previous versions.
Some devices are designed to deny users access to install or run
modified versions of the software inside them, although the manufacturer
can do so. This is fundamentally incompatible with the aim of
protecting users' freedom to change the software. The systematic
pattern of such abuse occurs in the area of products for individuals to
use, which is precisely where it is most unacceptable. Therefore, we
have designed this version of the GPL to prohibit the practice for those
products. If such problems arise substantially in other domains, we
stand ready to extend this provision to those domains in future versions
of the GPL, as needed to protect the freedom of users.
Finally, every program is threatened constantly by software patents.
States should not allow patents to restrict development and use of
software on general-purpose computers, but in those that do, we wish to
avoid the special danger that patents applied to a free program could
make it effectively proprietary. To prevent this, the GPL assures that
patents cannot be used to render the program non-free.
The precise terms and conditions for copying, distribution and
modification follow.
TERMS AND CONDITIONS
0. Definitions.
"This License" refers to version 3 of the GNU General Public License.
"Copyright" also means copyright-like laws that apply to other kinds of
works, such as semiconductor masks.
"The Program" refers to any copyrightable work licensed under this
License. Each licensee is addressed as "you". "Licensees" and
"recipients" may be individuals or organizations.
To "modify" a work means to copy from or adapt all or part of the work
in a fashion requiring copyright permission, other than the making of an
exact copy. The resulting work is called a "modified version" of the
earlier work or a work "based on" the earlier work.
A "covered work" means either the unmodified Program or a work based
on the Program.
To "propagate" a work means to do anything with it that, without
permission, would make you directly or secondarily liable for
infringement under applicable copyright law, except executing it on a
computer or modifying a private copy. Propagation includes copying,
distribution (with or without modification), making available to the
public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other
parties to make or receive copies. Mere interaction with a user through
a computer network, with no transfer of a copy, is not conveying.
An interactive user interface displays "Appropriate Legal Notices"
to the extent that it includes a convenient and prominently visible
feature that (1) displays an appropriate copyright notice, and (2)
tells the user that there is no warranty for the work (except to the
extent that warranties are provided), that licensees may convey the
work under this License, and how to view a copy of this License. If
the interface presents a list of user commands or options, such as a
menu, a prominent item in the list meets this criterion.
1. Source Code.
The "source code" for a work means the preferred form of the work
for making modifications to it. "Object code" means any non-source
form of a work.
A "Standard Interface" means an interface that either is an official
standard defined by a recognized standards body, or, in the case of
interfaces specified for a particular programming language, one that
is widely used among developers working in that language.
The "System Libraries" of an executable work include anything, other
than the work as a whole, that (a) is included in the normal form of
packaging a Major Component, but which is not part of that Major
Component, and (b) serves only to enable use of the work with that
Major Component, or to implement a Standard Interface for which an
implementation is available to the public in source code form. A
"Major Component", in this context, means a major essential component
(kernel, window system, and so on) of the specific operating system
(if any) on which the executable work runs, or a compiler used to
produce the work, or an object code interpreter used to run it.
The "Corresponding Source" for a work in object code form means all
the source code needed to generate, install, and (for an executable
work) run the object code and to modify the work, including scripts to
control those activities. However, it does not include the work's
System Libraries, or general-purpose tools or generally available free
programs which are used unmodified in performing those activities but
which are not part of the work. For example, Corresponding Source
includes interface definition files associated with source files for
the work, and the source code for shared libraries and dynamically
linked subprograms that the work is specifically designed to require,
such as by intimate data communication or control flow between those
subprograms and other parts of the work.
The Corresponding Source need not include anything that users
can regenerate automatically from other parts of the Corresponding
Source.
The Corresponding Source for a work in source code form is that
same work.
2. Basic Permissions.
All rights granted under this License are granted for the term of
copyright on the Program, and are irrevocable provided the stated
conditions are met. This License explicitly affirms your unlimited
permission to run the unmodified Program. The output from running a
covered work is covered by this License only if the output, given its
content, constitutes a covered work. This License acknowledges your
rights of fair use or other equivalent, as provided by copyright law.
You may make, run and propagate covered works that you do not
convey, without conditions so long as your license otherwise remains
in force. You may convey covered works to others for the sole purpose
of having them make modifications exclusively for you, or provide you
with facilities for running those works, provided that you comply with
the terms of this License in conveying all material for which you do
not control copyright. Those thus making or running the covered works
for you must do so exclusively on your behalf, under your direction
and control, on terms that prohibit them from making any copies of
your copyrighted material outside their relationship with you.
Conveying under any other circumstances is permitted solely under
the conditions stated below. Sublicensing is not allowed; section 10
makes it unnecessary.
3. Protecting Users' Legal Rights From Anti-Circumvention Law.
No covered work shall be deemed part of an effective technological
measure under any applicable law fulfilling obligations under article
11 of the WIPO copyright treaty adopted on 20 December 1996, or
similar laws prohibiting or restricting circumvention of such
measures.
When you convey a covered work, you waive any legal power to forbid
circumvention of technological measures to the extent such circumvention
is effected by exercising rights under this License with respect to
the covered work, and you disclaim any intention to limit operation or
modification of the work as a means of enforcing, against the work's
users, your or third parties' legal rights to forbid circumvention of
technological measures.
4. Conveying Verbatim Copies.
You may convey verbatim copies of the Program's source code as you
receive it, in any medium, provided that you conspicuously and
appropriately publish on each copy an appropriate copyright notice;
keep intact all notices stating that this License and any
non-permissive terms added in accord with section 7 apply to the code;
keep intact all notices of the absence of any warranty; and give all
recipients a copy of this License along with the Program.
You may charge any price or no price for each copy that you convey,
and you may offer support or warranty protection for a fee.
5. Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to
produce it from the Program, in the form of source code under the
terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified
it, and giving a relevant date.
b) The work must carry prominent notices stating that it is
released under this License and any conditions added under section
7. This requirement modifies the requirement in section 4 to
"keep intact all notices".
c) You must license the entire work, as a whole, under this
License to anyone who comes into possession of a copy. This
License will therefore apply, along with any applicable section 7
additional terms, to the whole of the work, and all its parts,
regardless of how they are packaged. This License gives no
permission to license the work in any other way, but it does not
invalidate such permission if you have separately received it.
d) If the work has interactive user interfaces, each must display
Appropriate Legal Notices; however, if the Program has interactive
interfaces that do not display Appropriate Legal Notices, your
work need not make them do so.
A compilation of a covered work with other separate and independent
works, which are not by their nature extensions of the covered work,
and which are not combined with it such as to form a larger program,
in or on a volume of a storage or distribution medium, is called an
"aggregate" if the compilation and its resulting copyright are not
used to limit the access or legal rights of the compilation's users
beyond what the individual works permit. Inclusion of a covered work
in an aggregate does not cause this License to apply to the other
parts of the aggregate.
6. Conveying Non-Source Forms.
You may convey a covered work in object code form under the terms
of sections 4 and 5, provided that you also convey the
machine-readable Corresponding Source under the terms of this License,
in one of these ways:
a) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by the
Corresponding Source fixed on a durable physical medium
customarily used for software interchange.
b) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by a
written offer, valid for at least three years and valid for as
long as you offer spare parts or customer support for that product
model, to give anyone who possesses the object code either (1) a
copy of the Corresponding Source for all the software in the
product that is covered by this License, on a durable physical
medium customarily used for software interchange, for a price no
more than your reasonable cost of physically performing this
conveying of source, or (2) access to copy the
Corresponding Source from a network server at no charge.
c) Convey individual copies of the object code with a copy of the
written offer to provide the Corresponding Source. This
alternative is allowed only occasionally and noncommercially, and
only if you received the object code with such an offer, in accord
with subsection 6b.
d) Convey the object code by offering access from a designated
place (gratis or for a charge), and offer equivalent access to the
Corresponding Source in the same way through the same place at no
further charge. You need not require recipients to copy the
Corresponding Source along with the object code. If the place to
copy the object code is a network server, the Corresponding Source
may be on a different server (operated by you or a third party)
that supports equivalent copying facilities, provided you maintain
clear directions next to the object code saying where to find the
Corresponding Source. Regardless of what server hosts the
Corresponding Source, you remain obligated to ensure that it is
available for as long as needed to satisfy these requirements.
e) Convey the object code using peer-to-peer transmission, provided
you inform other peers where the object code and Corresponding
Source of the work are being offered to the general public at no
charge under subsection 6d.
A separable portion of the object code, whose source code is excluded
from the Corresponding Source as a System Library, need not be
included in conveying the object code work.
A "User Product" is either (1) a "consumer product", which means any
tangible personal property which is normally used for personal, family,
or household purposes, or (2) anything designed or sold for incorporation
into a dwelling. In determining whether a product is a consumer product,
doubtful cases shall be resolved in favor of coverage. For a particular
product received by a particular user, "normally used" refers to a
typical or common use of that class of product, regardless of the status
of the particular user or of the way in which the particular user
actually uses, or expects or is expected to use, the product. A product
is a consumer product regardless of whether the product has substantial
commercial, industrial or non-consumer uses, unless such uses represent
the only significant mode of use of the product.
"Installation Information" for a User Product means any methods,
procedures, authorization keys, or other information required to install
and execute modified versions of a covered work in that User Product from
a modified version of its Corresponding Source. The information must
suffice to ensure that the continued functioning of the modified object
code is in no case prevented or interfered with solely because
modification has been made.
If you convey an object code work under this section in, or with, or
specifically for use in, a User Product, and the conveying occurs as
part of a transaction in which the right of possession and use of the
User Product is transferred to the recipient in perpetuity or for a
fixed term (regardless of how the transaction is characterized), the
Corresponding Source conveyed under this section must be accompanied
by the Installation Information. But this requirement does not apply
if neither you nor any third party retains the ability to install
modified object code on the User Product (for example, the work has
been installed in ROM).
The requirement to provide Installation Information does not include a
requirement to continue to provide support service, warranty, or updates
for a work that has been modified or installed by the recipient, or for
the User Product in which it has been modified or installed. Access to a
network may be denied when the modification itself materially and
adversely affects the operation of the network or violates the rules and
protocols for communication across the network.
Corresponding Source conveyed, and Installation Information provided,
in accord with this section must be in a format that is publicly
documented (and with an implementation available to the public in
source code form), and must require no special password or key for
unpacking, reading or copying.
7. Additional Terms.
"Additional permissions" are terms that supplement the terms of this
License by making exceptions from one or more of its conditions.
Additional permissions that are applicable to the entire Program shall
be treated as though they were included in this License, to the extent
that they are valid under applicable law. If additional permissions
apply only to part of the Program, that part may be used separately
under those permissions, but the entire Program remains governed by
this License without regard to the additional permissions.
When you convey a copy of a covered work, you may at your option
remove any additional permissions from that copy, or from any part of
it. (Additional permissions may be written to require their own
removal in certain cases when you modify the work.) You may place
additional permissions on material, added by you to a covered work,
for which you have or can give appropriate copyright permission.
Notwithstanding any other provision of this License, for material you
add to a covered work, you may (if authorized by the copyright holders of
that material) supplement the terms of this License with terms:
a) Disclaiming warranty or limiting liability differently from the
terms of sections 15 and 16 of this License; or
b) Requiring preservation of specified reasonable legal notices or
author attributions in that material or in the Appropriate Legal
Notices displayed by works containing it; or
c) Prohibiting misrepresentation of the origin of that material, or
requiring that modified versions of such material be marked in
reasonable ways as different from the original version; or
d) Limiting the use for publicity purposes of names of licensors or
authors of the material; or
e) Declining to grant rights under trademark law for use of some
trade names, trademarks, or service marks; or
f) Requiring indemnification of licensors and authors of that
material by anyone who conveys the material (or modified versions of
it) with contractual assumptions of liability to the recipient, for
any liability that these contractual assumptions directly impose on
those licensors and authors.
All other non-permissive additional terms are considered "further
restrictions" within the meaning of section 10. If the Program as you
received it, or any part of it, contains a notice stating that it is
governed by this License along with a term that is a further
restriction, you may remove that term. If a license document contains
a further restriction but permits relicensing or conveying under this
License, you may add to a covered work material governed by the terms
of that license document, provided that the further restriction does
not survive such relicensing or conveying.
If you add terms to a covered work in accord with this section, you
must place, in the relevant source files, a statement of the
additional terms that apply to those files, or a notice indicating
where to find the applicable terms.
Additional terms, permissive or non-permissive, may be stated in the
form of a separately written license, or stated as exceptions;
the above requirements apply either way.
8. Termination.
You may not propagate or modify a covered work except as expressly
provided under this License. Any attempt otherwise to propagate or
modify it is void, and will automatically terminate your rights under
this License (including any patent licenses granted under the third
paragraph of section 11).
However, if you cease all violation of this License, then your
license from a particular copyright holder is reinstated (a)
provisionally, unless and until the copyright holder explicitly and
finally terminates your license, and (b) permanently, if the copyright
holder fails to notify you of the violation by some reasonable means
prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is
reinstated permanently if the copyright holder notifies you of the
violation by some reasonable means, this is the first time you have
received notice of violation of this License (for any work) from that
copyright holder, and you cure the violation prior to 30 days after
your receipt of the notice.
Termination of your rights under this section does not terminate the
licenses of parties who have received copies or rights from you under
this License. If your rights have been terminated and not permanently
reinstated, you do not qualify to receive new licenses for the same
material under section 10.
9. Acceptance Not Required for Having Copies.
You are not required to accept this License in order to receive or
run a copy of the Program. Ancillary propagation of a covered work
occurring solely as a consequence of using peer-to-peer transmission
to receive a copy likewise does not require acceptance. However,
nothing other than this License grants you permission to propagate or
modify any covered work. These actions infringe copyright if you do
not accept this License. Therefore, by modifying or propagating a
covered work, you indicate your acceptance of this License to do so.
10. Automatic Licensing of Downstream Recipients.
Each time you convey a covered work, the recipient automatically
receives a license from the original licensors, to run, modify and
propagate that work, subject to this License. You are not responsible
for enforcing compliance by third parties with this License.
An "entity transaction" is a transaction transferring control of an
organization, or substantially all assets of one, or subdividing an
organization, or merging organizations. If propagation of a covered
work results from an entity transaction, each party to that
transaction who receives a copy of the work also receives whatever
licenses to the work the party's predecessor in interest had or could
give under the previous paragraph, plus a right to possession of the
Corresponding Source of the work from the predecessor in interest, if
the predecessor has it or can get it with reasonable efforts.
You may not impose any further restrictions on the exercise of the
rights granted or affirmed under this License. For example, you may
not impose a license fee, royalty, or other charge for exercise of
rights granted under this License, and you may not initiate litigation
(including a cross-claim or counterclaim in a lawsuit) alleging that
any patent claim is infringed by making, using, selling, offering for
sale, or importing the Program or any portion of it.
11. Patents.
A "contributor" is a copyright holder who authorizes use under this
License of the Program or a work on which the Program is based. The
work thus licensed is called the contributor's "contributor version".
A contributor's "essential patent claims" are all patent claims
owned or controlled by the contributor, whether already acquired or
hereafter acquired, that would be infringed by some manner, permitted
by this License, of making, using, or selling its contributor version,
but do not include claims that would be infringed only as a
consequence of further modification of the contributor version. For
purposes of this definition, "control" includes the right to grant
patent sublicenses in a manner consistent with the requirements of
this License.
Each contributor grants you a non-exclusive, worldwide, royalty-free
patent license under the contributor's essential patent claims, to
make, use, sell, offer for sale, import and otherwise run, modify and
propagate the contents of its contributor version.
In the following three paragraphs, a "patent license" is any express
agreement or commitment, however denominated, not to enforce a patent
(such as an express permission to practice a patent or covenant not to
sue for patent infringement). To "grant" such a patent license to a
party means to make such an agreement or commitment not to enforce a
patent against the party.
If you convey a covered work, knowingly relying on a patent license,
and the Corresponding Source of the work is not available for anyone
to copy, free of charge and under the terms of this License, through a
publicly available network server or other readily accessible means,
then you must either (1) cause the Corresponding Source to be so
available, or (2) arrange to deprive yourself of the benefit of the
patent license for this particular work, or (3) arrange, in a manner
consistent with the requirements of this License, to extend the patent
license to downstream recipients. "Knowingly relying" means you have
actual knowledge that, but for the patent license, your conveying the
covered work in a country, or your recipient's use of the covered work
in a country, would infringe one or more identifiable patents in that
country that you have reason to believe are valid.
If, pursuant to or in connection with a single transaction or
arrangement, you convey, or propagate by procuring conveyance of, a
covered work, and grant a patent license to some of the parties
receiving the covered work authorizing them to use, propagate, modify
or convey a specific copy of the covered work, then the patent license
you grant is automatically extended to all recipients of the covered
work and works based on it.
A patent license is "discriminatory" if it does not include within
the scope of its coverage, prohibits the exercise of, or is
conditioned on the non-exercise of one or more of the rights that are
specifically granted under this License. You may not convey a covered
work if you are a party to an arrangement with a third party that is
in the business of distributing software, under which you make payment
to the third party based on the extent of your activity of conveying
the work, and under which the third party grants, to any of the
parties who would receive the covered work from you, a discriminatory
patent license (a) in connection with copies of the covered work
conveyed by you (or copies made from those copies), or (b) primarily
for and in connection with specific products or compilations that
contain the covered work, unless you entered into that arrangement,
or that patent license was granted, prior to 28 March 2007.
Nothing in this License shall be construed as excluding or limiting
any implied license or other defenses to infringement that may
otherwise be available to you under applicable patent law.
12. No Surrender of Others' Freedom.
If conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot convey a
covered work so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you may
not convey it at all. For example, if you agree to terms that obligate you
to collect a royalty for further conveying from those to whom you convey
the Program, the only way you could satisfy both those terms and this
License would be to refrain entirely from conveying the Program.
13. Use with the GNU Affero General Public License.
Notwithstanding any other provision of this License, you have
permission to link or combine any covered work with a work licensed
under version 3 of the GNU Affero General Public License into a single
combined work, and to convey the resulting work. The terms of this
License will continue to apply to the part which is the covered work,
but the special requirements of the GNU Affero General Public License,
section 13, concerning interaction through a network will apply to the
combination as such.
14. Revised Versions of this License.
The Free Software Foundation may publish revised and/or new versions of
the GNU General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the
Program specifies that a certain numbered version of the GNU General
Public License "or any later version" applies to it, you have the
option of following the terms and conditions either of that numbered
version or of any later version published by the Free Software
Foundation. If the Program does not specify a version number of the
GNU General Public License, you may choose any version ever published
by the Free Software Foundation.
If the Program specifies that a proxy can decide which future
versions of the GNU General Public License can be used, that proxy's
public statement of acceptance of a version permanently authorizes you
to choose that version for the Program.
Later license versions may give you additional or different
permissions. However, no additional obligations are imposed on any
author or copyright holder as a result of your choosing to follow a
later version.
15. Disclaimer of Warranty.
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
16. Limitation of Liability.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.
17. Interpretation of Sections 15 and 16.
If the disclaimer of warranty and limitation of liability provided
above cannot be given local legal effect according to their terms,
reviewing courts shall apply local law that most closely approximates
an absolute waiver of all civil liability in connection with the
Program, unless a warranty or assumption of liability accompanies a
copy of the Program in return for a fee.
END OF TERMS AND CONDITIONS
How to Apply These Terms to Your New Programs
If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
state the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
Also add information on how to contact you by electronic and paper mail.
If the program does terminal interaction, make it output a short
notice like this when it starts in an interactive mode:
<program> Copyright (C) <year> <name of author>
This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, your program's commands
might be different; for a GUI interface, you would use an "about box".
You should also get your employer (if you work as a programmer) or school,
if any, to sign a "copyright disclaimer" for the program, if necessary.
For more information on this, and how to apply and follow the GNU GPL, see
<https://www.gnu.org/licenses/>.
The GNU General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you
may consider it more useful to permit linking proprietary applications with
the library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License. But first, please read
<https://www.gnu.org/licenses/why-not-lgpl.html>.
README.md
+183 -244
View File
@@ -1,102 +1,19 @@
# dejima-ci-hardener
Hardens a CI runner that executes agent-authored code on every pull request,
and gives it as little as possible to execute it with: no Docker socket in job
containers, and deny-by-default egress whose allowlist comes from
`.dejima/allowed-domains` on each served repository's `main`.
Hardens a CI runner host that executes agent-authored code on every pull
request, and gives its jobs as little as possible to execute it with.
**This is the only thing you clone onto a CI box.** It is self-contained -
there is no step that copies files out of a dejima checkout, and dejima itself
never needs to be present here.
There are two kinds of runner host, and which one you are building changes
almost everything below. Pick the tier first, then follow only its half.
For *why* any of this is necessary - what a CI run adds over a sandbox, and
why no approval gate is offered - see dejima's `docs/ci.md`. You do not need
it to follow this.
## The runner host itself
`install.sh` hardens the runners on a machine you already have. For the
privileged tier that is the wrong starting point - its jobs get the Docker
socket, so the assumption has to be that a compromised job gets root on that
host, and handing an existing machine to that assumption is not a decision
many people would make deliberately.
**[proxmox/README.md](proxmox/README.md)** builds one instead: a runbook for
the Proxmox web console, then one script inside the VM.
```sh
# inside the new VM, once the console steps are done
sudo ./bin/dejima-ci-bootstrap --role runtime --mgmt <your-ip>
```
On `--role runtime` that deliberately stops short of running `install.sh`. The
job network, its resolver and its DOCKER-USER rules all work by sitting in
front of job traffic on a bridge, and this tier's jobs are in the host's
network namespace where no bridge is. Installing them would cost an evening
and leave a host that looks like it has an egress allowlist.
The node half is a runbook rather than a script on purpose. Something that
runs `qm create` and writes into `/etc/pve/firewall/` is exactly what gets
pasted into a root shell unread, and the blast radius is every VM on the node.
The in-VM half is a script, because running that one blind costs you a machine
built to be lost.
### The boundary is on the node, not in the VM
The privileged tier hands job containers the Docker socket, which is
root-equivalent on their host. So the working assumption is that a compromised
job **has root in that VM** — and root can flush any firewall it can see.
Everything installed inside the VM is therefore defence in depth: real against
a compromised job that has not escaped its container, which is the ordinary
case, and worth nothing against one that has.
What holds either way is the Proxmox firewall, applied to the VM's tap device
and unreachable from the guest. It denies the VM every RFC1918 destination
except two pinholes — the forge on 443 and a resolver on 53 — and those two
are the entire remaining LAN exposure.
Public egress is deliberately *not* restricted there - not by address and not
by port. An address allowlist on the node would break on every CDN rotation,
and a port allowlist buys nothing: an earlier version permitted only tcp/80,
tcp/443 and udp/123, which broke apt on a first install, and anyone who owns
that box tunnels over 443 like everybody else. Neither would hold against VM
root in any case.
Note what that leaves. On this tier there is no egress allowlist in front of
job containers either - they are in the host's network namespace, so nothing
is in front of them. A privileged job reaches the whole public internet by
design. What it does not reach is your LAN.
The guarantee is narrower and honest, and it is the one worth making: an
attacker who owns that VM reaches the internet, and not your network.
One thing to check rather than assume, because its absence looks exactly like
it working: **the datacenter firewall must be enabled** (Datacenter → Firewall
→ Options → Firewall: Yes), and the VM's network device must have its own
**Firewall** checkbox ticked. Without either, the per-VM rules apply to
nothing. `pve-firewall compile` on the node is the check, and
[proxmox/README.md](proxmox/README.md) has a from-inside-the-VM check that
does not rely on reading configuration at all.
### Snapshot, and roll back on a cadence
The VM is worth losing, which is only true if losing it is cheap. Register the
runner, then snapshot — **VM → Snapshots → Take Snapshot**, named `clean`,
**before the first job runs** and with *Include RAM* unticked. One taken later
preserves whatever that job left; one taken before registration drops the
runner off the forge on every rollback; one with RAM restores the guest's clock
along with it, and a VM that thinks it is last month cannot verify a
certificate.
Rolling back after anything suspicious is not optional. Rolling back *on a
schedule* is the more valuable of the two, because it does not depend on your
having noticed — and it is the one that stops happening once doing it by hand
gets old. Proxmox schedules backups, not rollbacks, so it is a cron entry on
the node; [proxmox/README.md](proxmox/README.md) §6 has the line, why each part
of it is what it is, how to confirm it is still running, and why the baseline
needs refreshing every month or two.
## Do not install this next to dejima
This runs on a machine whose job is to execute untrusted code. dejima runs on
@@ -105,7 +22,65 @@ makes either worth anything, and this program holds a forge token and rewrites
a resolver's configuration - neither of which belongs within reach of a
sandbox session.
## Install
## Two tiers, and you must pick one
The difference is not how much a job is granted. It is **where the security
boundary sits**, and therefore what you have to build to have one at all.
| | **confined** | **privileged** |
| --- | --- | --- |
| Built by | `--role confined` | `--role privileged` |
| Docker socket in job containers | no | **yes** |
| Job networking | its own bridge | the host's namespace |
| Egress allowlist in front of jobs | yes, from each repo's `main` | **none, by design** |
| The boundary is | in the host | on the hypervisor |
| Treat this host as | expected to hold | already lost |
| Runs `install.sh` | yes | no, deliberately |
| For | tests, lint, anything needing no daemon | image builds, tests needing a real daemon or host networking |
**Confined** is where everything belongs that can go there. Its jobs get no
Docker socket, run on a bridge this tool creates, and reach only the domains
the served repositories declare in `.dejima/allowed-domains` on their `main`.
`install.sh` builds all of that, and it is reasonable to run on a machine you
already have.
**Privileged** exists for the jobs that cannot work otherwise: building a
Docker image, or dejima's own runtime tests, which assert things about a real
daemon and a real host network. Both grants are root-equivalent on the host,
and neither can be reduced without breaking the job they exist for. That tier
needs a machine built to be lost, not a machine you own.
### The privileged tier has no egress allowlist, on purpose
This is the split's most important consequence and the easiest to read as an
omission, so plainly: it is not one.
The confined tier's allowlist works because its jobs cannot reach it. The
rules sit on a bridge their traffic has to cross, in a network namespace they
are not in, enforced by a kernel they hold no privileged access to.
A privileged job has all three. It holds the Docker socket, which is
root-equivalent on that host, and it runs in the host's own network namespace.
Any filter installed there is one it can flush in a single command. Installing
it anyway would cost an evening and leave a host that *looks* like it has an
egress allowlist, which is worse than one that plainly does not: the next
person to reason about that machine reasons from the label.
So `dejima-ci-bootstrap --role privileged` deliberately stops short of running
`install.sh`, and the boundary moves outward - to the hypervisor's firewall,
applied to the VM's tap device, where the guest cannot reach it.
The guarantee that survives is narrower and honest, and it is the one worth
making: **an attacker who owns that VM reaches the internet, and not your
network.**
---
# The confined tier
## Build or harden the host
Either works. On a machine you already have:
```sh
sudo apt-get install ipset iptables dnsmasq
@@ -114,11 +89,21 @@ cd dejima-ci-hardener
sudo ./install.sh
```
`install.sh` prompts for each value, offering whatever is already configured
as the default, and does the whole job: creates the Docker network with a
fixed bridge name, writes the resolver config, installs
`dejima-ci-allowlist` and `dejima-ci-firewall`, installs and enables the
systemd units, and renders the first allowlist.
On a fresh VM, [proxmox/README.md](proxmox/README.md) builds one and
`dejima-ci-bootstrap` finishes it - packages, sshd, unattended upgrades, the
host's own egress policy, and then `install.sh` for the job allowlist:
```sh
sudo ./bin/dejima-ci-bootstrap --role confined --mgmt <your-ip>
```
## What `install.sh` does
It prompts for each value, offering whatever is already configured as the
default, and does the whole job: creates the Docker network with a fixed
bridge name, writes the resolver config, installs `dejima-ci-allowlist` and
`dejima-ci-firewall`, installs and enables the systemd units, and renders the
first allowlist.
**Idempotent** - re-running and pressing Enter at every prompt changes
nothing, so it is also how to change one setting later.
@@ -135,7 +120,7 @@ What it asks for:
| Repositories | `owner/repo` per line - adding one trusts that repository's `main` |
| Base domains | allowed regardless of any project declaration |
### The upstream resolver
## The upstream resolver
The one setting that is genuinely ambiguous, so: it is the resolver **dnsmasq
forwards allowed queries to** - the one that actually knows the answers.
@@ -158,31 +143,110 @@ if that name is internal, this has to be your internal DNS. `no-resolv` in the
dnsmasq config means it will not read `/etc/resolv.conf` itself, which is why
it must be named explicitly.
### The runners
## The runner
`install.sh` does not register runners - that needs a token from your forge
and is the one step it cannot do for you. `examples/` has the configuration
they need:
and is the one step it cannot do for you.
| File | Install as |
| --- | --- |
| `examples/act-runner-ci.yaml` | the ordinary runner's `config.yaml` |
| `examples/act-runner-runtime.yaml` | the privileged runner's, on its own host |
| `examples/docker-compose.yml` | several ordinary runners on one host |
| `examples/docker-compose-runtime.yml` | the privileged one |
| `examples/act-runner-confined.yaml` | the runner's `config.yaml` |
| `examples/docker-compose-confined.yml` | several runners on one host |
| `examples/docker-daemon.json` | `/etc/docker/daemon.json` - merge, do not overwrite |
Set `--dns=<your gateway>` in the ordinary runner's `container.options`, and
register at **repository or organization** scope. That last part is not
housekeeping — see below.
Set `--dns=<your gateway>` in `container.options`, and register at
**repository or organization** scope - see [Register scope is the
boundary](#register-scope-is-the-boundary-the-label-is-not).
### Register scope is the boundary; the label is not
---
# The privileged tier
## Build the VM. Do not harden an existing one.
`install.sh` hardens the runners on a machine you already have. For this tier
that is the wrong starting point - its jobs get the Docker socket, so the
assumption has to be that a compromised job gets root on that host, and
handing an existing machine to that assumption is not a decision many people
would make deliberately.
**[proxmox/README.md](proxmox/README.md)** builds one instead: a runbook for
the Proxmox web console, then one script inside the VM.
```sh
# inside the new VM, once the console steps are done
sudo ./bin/dejima-ci-bootstrap --role privileged --mgmt <your-ip>
```
The node half is a runbook rather than a script on purpose. Something that
runs `qm create` and writes into `/etc/pve/firewall/` is exactly what gets
pasted into a root shell unread, and the blast radius is every VM on the node.
The in-VM half is a script, because running that one blind costs you a machine
built to be lost.
## The boundary is on the node, not in the VM
Everything installed inside the VM is defence in depth: real against a
compromised job that has not escaped its container, which is the ordinary
case, and worth nothing against one that has.
What holds either way is the Proxmox firewall, applied to the VM's tap device
and unreachable from the guest. It denies the VM every RFC1918 destination
except two pinholes - the forge on 443 and a resolver on 53 - and those two
are the entire remaining LAN exposure. Public egress is deliberately *not*
restricted there;
[docs/reference.md](docs/reference.md#why-public-egress-is-not-restricted-on-the-node)
says why, and why a port allowlist there bought nothing.
One thing to check rather than assume, because its absence looks exactly like
it working: **the datacenter firewall must be enabled** (Datacenter → Firewall
→ Options → Firewall: Yes), and the VM's network device must have its own
**Firewall** checkbox ticked. Without either, the per-VM rules apply to
nothing. `pve-firewall compile` on the node is the check, and
[proxmox/README.md](proxmox/README.md) has a from-inside-the-VM check that
does not rely on reading configuration at all.
## Snapshot, and roll back on a cadence
The VM is worth losing, which is only true if losing it is cheap. Register the
runner, then snapshot - **VM → Snapshots → Take Snapshot**, named `clean`,
**before the first job runs** and with *Include RAM* unticked. One taken later
preserves whatever that job left; one taken before registration drops the
runner off the forge on every rollback; one with RAM restores the guest's clock
along with it, and a VM that thinks it is last month cannot verify a
certificate.
Rolling back after anything suspicious is not optional. Rolling back *on a
schedule* is the more valuable of the two, because it does not depend on your
having noticed - and it is the one that stops happening once doing it by hand
gets old. Proxmox schedules backups, not rollbacks, so it is a cron entry on
the node; [proxmox/README.md](proxmox/README.md) §6 has the line, why each part
of it is what it is, how to confirm it is still running, and why the baseline
needs refreshing every month or two.
## The runner
| File | Install as |
| --- | --- |
| `examples/act-runner-privileged.yaml` | the runner's `config.yaml`, on its own host |
| `examples/docker-compose-privileged.yml` | the runner itself |
Give it one label that nothing else carries, and grant it to as few
repositories as you can stand. There is no `--dns` to set and no job network
to join: its jobs are in the host's namespace, which is the grant the tier
exists for.
---
# Both tiers
## Register scope is the boundary; the label is not
A dedicated label routes honest jobs to the runner you prepared. It stops
nothing else, because `runs-on:` lives in the workflow file and the workflow
file lives in the branch. A branch can name any label it likes, so a session
that changes `runs-on: dejima-ci` to `runs-on: ubuntu-latest` and opens a pull
request has stepped outside everything this tool installs — provided some
request has stepped outside everything this tool installs - provided some
runner answering that label will take the job.
Whether one will depends on where your runners are registered:
@@ -222,6 +286,9 @@ egress checks fail, the resolver is fine and the packet filter is not.
## Use
Confined-tier commands. A privileged host installs none of them, which is the
point of that tier rather than a gap in it.
```sh
dejima-ci-allowlist check # what would change, without changing it
dejima-ci-allowlist render # fetch, validate, install, reload if changed
@@ -274,143 +341,15 @@ bats test/
Runs against a stubbed `curl` and `dnsmasq`.
---
## Further reading
# Reference
| | |
| --- | --- |
| [docs/](docs/) | Index of everything below |
| [docs/reference.md](docs/reference.md) | The two `docker.sock` settings, `CONFIG_FILE`, the `/data` volume, compose networks against job networks, sharing a job network between runners, why the privileged runner is separate, why the resolver is per-runner, the cache server |
| [proxmox/README.md](proxmox/README.md) | Building the sacrificial VM: node firewall rules, bootstrap, snapshot, scheduled rollback |
| dejima's `docs/ci.md` | Why a CI runner needs any of this, and why there is no approval gate |
Everything below is background. You do not need it to complete the setup.
## License
## The two settings named `docker.sock`
The compose bind and `container.docker_host` name the same path and do
opposite jobs. This is the easiest thing here to conflate:
| | Effect | Needed? |
| --- | --- | --- |
| The compose bind | The runner can reach the daemon | Yes — without it nothing runs |
| `docker_host: "-"` | Job containers *cannot* | Yes — without it every job is root on this host |
Setting the second does not remove the first, and removing the first does not
make jobs safe — it just stops the runner working.
Whether a job gets the socket is a property of **the runner**, not of the job:
a job asking for nothing unusual still receives whatever its runner grants
everything it runs.
## `CONFIG_FILE`, and why an edit can appear to do nothing
Without `CONFIG_FILE=/config.yaml` in the environment, `act_runner` does not
read the config file **at all** — it starts happily on built-in defaults,
which means the socket goes into every job container and nothing in any log
says so. If a config change appears to have no effect, check this first.
Mount the config read-only: a runner has no reason to rewrite the file that
decides what its jobs are granted.
## The `/data` volume
It holds `.runner`, written at registration: the runner's identity and its
credential for your forge (`runner.file: .runner`, resolved against the
container's working directory, which is `/data` in the official image).
- **Keep it** and `--force-recreate`, an image bump or a reboot brings back
*the same registered runner*. Without it the container comes up with no
identity — with a token still in the environment it silently registers a
*new* runner, leaving a dead entry in your forge's runner list; without one
it fails to start.
- **Delete it deliberately** to re-register: a runner moved to another
instance, or a label set you would rather establish fresh.
- **It is the only state.** Nothing else in the container is worth preserving,
which is what lets the rest of this treat a runner as disposable.
Labels come from `runner.labels` on every start rather than being frozen at
registration, so changing one is an edit and a restart.
## Compose networks and the job network are different worlds
Running several runners from one compose file makes them share a network — and
that is not the network worth worrying about.
Every service in that file is an `act_runner` *process*. The containers that
execute agent-authored code are created by the *daemon*, on the network named
by `container.network`, and compose knows nothing about them. Attaching a
service to a compose network does not put its jobs there.
So the shared network holds runner processes and nothing else, and those are
mutually trusted — each already holds the Docker socket, so separating them
protects nothing. The example names one `runners` network explicitly rather
than relying on the implicit `default`, so nothing joins it later by accident.
The compose file deliberately does **not** attach a runner to the job network
(a process holding the socket has no business sharing a network with untrusted
code), and does **not** put the runner behind the egress allowlist (those rules
filter the job bridge only — the runner is trusted, the jobs are not).
## Several runners, and whether they should share a job network
What *is* shared, with one compose file or five, is the job network: every
runner whose config names `dejima-ci` puts its jobs on that bridge. With N
runners at `capacity: 1` you get up to N jobs there at once, able to reach each
other and each other's service containers.
- **One shared job network** — the default, and why several runners can share
one config file. Every job on it has the same egress allowlist and none can
reach the host or the runners, so the added exposure is strictly job-to-job.
If the concern is a prompt-injected session reaching *outward*, this already
covers it.
- **One per runner** — worth it if concurrent jobs come from repositories that
should not interfere, or if service containers hold anything a neighbour
should not see. Costs a config file and a `docker network create` each, and
the firewall rules looped over each bridge.
The `ci-allow` ipset is shared either way — it holds destinations, not
per-bridge state, so one set and one dnsmasq serve every bridge. The wrinkle
is DNS: `daemon.json`'s `dns` is a single daemon-wide address that would sit
on one bridge and be unreachable from the others. Give dnsmasq a stable
address on a dummy interface, point `daemon.json` there, and let each
bridge's `INPUT` rule allow port 53 to it.
## Why the privileged runner is separate
It is handed the Docker daemon and the host's network namespace, because one
job cannot work without them and neither grant can be reduced without breaking
it. A job on it can take the machine it runs on — so that machine should exist
for nothing else. A separate compose file makes a separate host the default,
rather than something to disentangle later.
Where the two tiers do share a machine, be clear about what that is: **label
tiering, not host separation.** A job denied the socket still shares a kernel
with a runner that has it. A real reduction over granting everything to
everything; not isolation between the tiers.
It needs no bind for `/tmp/dejima-ci` — the job container gets that path
because the workflow asks for it and `container.valid_volumes` permits it, and
the host daemon resolves the source host-side.
## Why the resolver is set per-runner, not daemon-wide
`daemon.json` has a `dns` key, and pointing it at the allowlisting resolver is
the obvious way to make job containers use it. It is also wrong here, and the
failure is not obvious.
That setting applies to **every container the daemon starts**. On a host that
also runs the privileged tier, that includes containers created by its jobs
over the Docker socket — which are not job containers and were never meant to
be behind the CI resolver.
A dejima sandbox is exactly such a container, and it has a firewall of its
own. It allowlists precisely one host address: its own bridge gateway, taken
from its default route. A resolver on a *different* bridge is therefore
rejected by the sandbox's own rules, so nothing inside it resolves — and the
symptom is not "DNS is misconfigured" but the sandbox's startup self-check
reporting that a host it allows is unreachable.
`--dns` in `container.options` scopes the resolver to the containers that
runner starts, which is the set that should have it.
## The cache server
`act_runner` enables it by default, and it binds on the host *specifically* so
job containers can reach it — an inbound service exposed to untrusted code. The
firewall here blocks exactly that. Leave it off, or add a rule knowing what you
are exposing.
GPL-3.0-or-later. See [LICENSE](LICENSE).
bin/dejima-ci-allowlist
+3
View File
@@ -1,4 +1,7 @@
#!/usr/bin/env bash
# SPDX-License-Identifier: GPL-3.0-or-later
# Copyright (C) 2026 Brad Anderson
#
# dejima-ci-allowlist - render a CI job network's DNS allowlist from the
# .dejima/allowed-domains declarations of the repositories this runner serves,
# read from the branch the operator controls.
bin/dejima-ci-bootstrap
+51 -38
View File
@@ -1,4 +1,7 @@
#!/usr/bin/env bash
# SPDX-License-Identifier: GPL-3.0-or-later
# Copyright (C) 2026 Brad Anderson
#
# Take a fresh Debian install to a hardened CI runner host.
#
# Runs INSIDE the VM. The firewall that keeps this machine off your LAN is not
@@ -22,16 +25,17 @@ SKIP_HARDENER=0
usage() {
cat >&2 <<EOF
Usage: sudo $PROG --role ci|runtime [--mgmt IP] [--skip-hardener]
Usage: sudo $PROG --role confined|privileged [--mgmt IP] [--skip-hardener]
--role ci the ordinary tier: no Docker socket in job containers,
deny-by-default job egress
--role runtime the privileged tier: its jobs get the socket and the host
network namespace. This host is the one to treat as lost
--mgmt IP restrict sshd to this source as well as the node firewall
--skip-hardener do the host hardening only, not the allowlist install.
Implied by --role runtime, where the allowlist has nothing
to attach to
--role confined jobs are confined on this host: no Docker socket in job
containers, own bridge network, deny-by-default egress
--role privileged jobs get the socket and the host network namespace, so
nothing on this host can confine them. Treat it as lost;
the boundary is the hypervisor's firewall
--mgmt IP restrict sshd to this source as well as the node firewall
--skip-hardener do the host hardening only, not the allowlist install.
Implied by --role privileged, where the allowlist has
nothing to attach to
EOF
exit 2
}
@@ -46,7 +50,16 @@ while [ $# -gt 0 ]; do
esac
shift
done
case "$ROLE" in ci|runtime) ;; *) usage ;; esac
# `ci` and `runtime` were the original spellings. They named the tiers after
# what dejima used them for rather than after what a job on them is granted,
# which is the thing that has to be legible to anyone else. Still accepted,
# and said out loud rather than translated silently.
case "$ROLE" in
confined|privileged) ;;
ci) echo "$PROG: --role ci is now --role confined" >&2; ROLE=confined ;;
runtime) echo "$PROG: --role runtime is now --role privileged" >&2; ROLE=privileged ;;
*) usage ;;
esac
die() { echo "$PROG: $*" >&2; exit 1; }
say() { printf '%s\n' "$*" >&2; }
@@ -70,14 +83,14 @@ apt-get update -qq
# machine whose value is a small surface - drops it. apt then reports success,
# the daemon is installed, and there is no `docker` binary at all.
#
# ipset and dnsmasq belong to install.sh, which only runs on the ordinary
# tier. The runtime tier's jobs are in the host network namespace, where a job
# network's resolver and its ipset have nothing to sit in front of - so on that
# role these install an unused listener on port 53, on the machine this same
# script strips listeners from forty lines below. iptables stays either way:
# the host's own egress policy uses it on both roles.
# ipset and dnsmasq belong to install.sh, which only runs on the confined
# tier. The privileged tier's jobs are in the host network namespace, where a
# job network's resolver and its ipset have nothing to sit in front of - so on
# that role these install an unused listener on port 53, on the machine this
# same script strips listeners from forty lines below. iptables stays either
# way: the host's own egress policy uses it on both roles.
tier_packages=""
if [ "$ROLE" = ci ]; then
if [ "$ROLE" = confined ]; then
tier_packages="ipset dnsmasq"
fi
apt-get install -y -qq --no-install-recommends \
@@ -193,11 +206,11 @@ fi
# particular announces the machine to the LAN it is meant to be isolated from.
say "Removing services a runner does not need..."
unwanted="avahi-daemon cups bluetooth modemmanager rpcbind"
# dnsmasq is not installed on a runtime host any more, but it was by earlier
# runs of this script, and an upgrade that leaves the old listener running is
# not an upgrade. Disabled rather than purged: someone co-locating both tiers
# on one machine still needs the package.
if [ "$ROLE" = runtime ]; then
# dnsmasq is not installed on a privileged host any more, but it was by
# earlier runs of this script, and an upgrade that leaves the old listener
# running is not an upgrade. Disabled rather than purged: someone co-locating
# both tiers on one machine still needs the package.
if [ "$ROLE" = privileged ]; then
unwanted="$unwanted dnsmasq"
fi
for svc in $unwanted; do
@@ -330,21 +343,21 @@ say "Enabling unattended security upgrades..."
dpkg-reconfigure -f noninteractive unattended-upgrades >/dev/null 2>&1 || true
# --- the hardener itself ----------------------------------------------------
# Only for the ordinary tier, and the reason is the same one this whole
# arrangement rests on: the runtime tier's jobs run in the HOST network
# Only for the confined tier, and the reason is the same one this whole
# arrangement rests on: the privileged tier's jobs run in the HOST network
# namespace. No bridge sits in front of them, so a job network, its resolver
# and its DOCKER-USER rules never see their traffic. Installing that machinery
# here would take a working evening and protect nothing, and - worse - would
# leave a runtime host that looks like it has an egress allowlist.
# leave a privileged host that looks like it has an egress allowlist.
#
# What protects the runtime host is the node's firewall. That is the whole
# What protects the privileged host is the node's firewall. That is the whole
# list.
if [ "$SKIP_HARDENER" -eq 1 ]; then
say
say "Skipping install.sh (--skip-hardener)."
elif [ "$ROLE" = runtime ]; then
elif [ "$ROLE" = privileged ]; then
say
say "Not running install.sh on a runtime host, deliberately."
say "Not running install.sh on a privileged host, deliberately."
say
say "Its jobs run in the host network namespace - that is the grant the tier"
say "exists for - so no bridge is in front of them and a job network, its"
@@ -355,7 +368,7 @@ elif [ "$ROLE" = runtime ]; then
say "What protects this host is the Proxmox node's firewall, and that is the"
say "entire list. See proxmox/README.md."
say
say "If this machine also runs ordinary runners - not recommended, and not"
say "If this machine also runs confined runners - not recommended, and not"
say "what the tiering is for - run ./install.sh by hand."
else
say
@@ -378,18 +391,18 @@ cat >&2 <<EOF
Enabling it before that file is filled in will cut this host off from the
forge, which is why it is not enabled for you.
On a runtime host this is the only egress control inside the VM, and it is
still not a boundary: root here can drop it. The node's firewall is the
On a privileged host this is the only egress control inside the VM, and it
is still not a boundary: root here can drop it. The node's firewall is the
boundary.
2. Register the runner. This script does not, because it needs a token from
your forge:
- role ci: label it something no other runner carries, and point
container.network / --dns at what install.sh printed
- role runtime: this is the privileged tier - socket and host netns.
Give it one label that nothing else carries, and grant
it to as few repositories as you can stand.
- role confined: label it something no other runner carries, and point
container.network / --dns at what install.sh printed
- role privileged: socket and host netns. Give it one label that nothing
else carries, and grant it to as few repositories as
you can stand.
3. In the Proxmox console, once step 2 shows the runner Online, snapshot this
VM - BEFORE the first job runs, and with 'Include RAM' UNTICKED:
@@ -412,6 +425,6 @@ cat >&2 <<EOF
Remember what is and is not enforced here. The node's firewall keeps this
machine off your LAN and cannot be reached from inside it. Everything this
script installed can be removed by root on this host - which, on the runtime
role, is what a compromised job is assumed to get.
script installed can be removed by root on this host - which, on the
privileged role, is what a compromised job is assumed to get.
EOF
bin/dejima-ci-firewall
BIN
View File
Binary file not shown.
docs/README.md
+15
View File
@@ -0,0 +1,15 @@
# Documentation
The [README](../README.md) is the whole of the setup. Everything here is
background, deeper explanation, or the reasoning behind a decision that looks
arbitrary from outside.
| | |
| --- | --- |
| [reference.md](reference.md) | The two settings named `docker.sock`, `CONFIG_FILE` and why an edit can appear to do nothing, the `/data` volume, compose networks against job networks, whether several runners should share a job network, why the privileged runner is separate, why public egress is not restricted on the node, why the resolver is set per-runner, the cache server |
| [../proxmox/README.md](../proxmox/README.md) | Building the sacrificial VM: node firewall rules and the order they go in, bootstrap, snapshotting clean, and the scheduled rollback |
| [../proxmox/firewall.fw.template](../proxmox/firewall.fw.template) | The same node firewall as one annotated file, easier to review in one piece than as a table. Reference, not something to install |
One document lives outside this repository: dejima's `docs/ci.md`, which
covers why a CI run needs any of this over a sandbox, and why no approval gate
is offered. You do not need it to follow the README.
docs/reference.md
+164
View File
@@ -0,0 +1,164 @@
# Reference
Background. You do not need any of it to complete the setup - the
[README](../README.md) is self-contained. This is where the reasoning lives
for the decisions that look arbitrary from outside, and the failure modes that
present as something other than what they are.
## The two settings named `docker.sock`
The compose bind and `container.docker_host` name the same path and do
opposite jobs. This is the easiest thing here to conflate:
| | Effect | Needed? |
| --- | --- | --- |
| The compose bind | The runner can reach the daemon | Yes — without it nothing runs |
| `docker_host: "-"` | Job containers *cannot* | Yes — without it every job is root on this host |
Setting the second does not remove the first, and removing the first does not
make jobs safe — it just stops the runner working.
Whether a job gets the socket is a property of **the runner**, not of the job:
a job asking for nothing unusual still receives whatever its runner grants
everything it runs.
## `CONFIG_FILE`, and why an edit can appear to do nothing
Without `CONFIG_FILE=/config.yaml` in the environment, `act_runner` does not
read the config file **at all** — it starts happily on built-in defaults,
which means the socket goes into every job container and nothing in any log
says so. If a config change appears to have no effect, check this first.
Mount the config read-only: a runner has no reason to rewrite the file that
decides what its jobs are granted.
## The `/data` volume
It holds `.runner`, written at registration: the runner's identity and its
credential for your forge (`runner.file: .runner`, resolved against the
container's working directory, which is `/data` in the official image).
- **Keep it** and `--force-recreate`, an image bump or a reboot brings back
*the same registered runner*. Without it the container comes up with no
identity — with a token still in the environment it silently registers a
*new* runner, leaving a dead entry in your forge's runner list; without one
it fails to start.
- **Delete it deliberately** to re-register: a runner moved to another
instance, or a label set you would rather establish fresh.
- **It is the only state.** Nothing else in the container is worth preserving,
which is what lets the rest of this treat a runner as disposable.
Labels come from `runner.labels` on every start rather than being frozen at
registration, so changing one is an edit and a restart.
## Compose networks and the job network are different worlds
Running several runners from one compose file makes them share a network — and
that is not the network worth worrying about.
Every service in that file is an `act_runner` *process*. The containers that
execute agent-authored code are created by the *daemon*, on the network named
by `container.network`, and compose knows nothing about them. Attaching a
service to a compose network does not put its jobs there.
So the shared network holds runner processes and nothing else, and those are
mutually trusted — each already holds the Docker socket, so separating them
protects nothing. The example names one `runners` network explicitly rather
than relying on the implicit `default`, so nothing joins it later by accident.
The compose file deliberately does **not** attach a runner to the job network
(a process holding the socket has no business sharing a network with untrusted
code), and does **not** put the runner behind the egress allowlist (those rules
filter the job bridge only — the runner is trusted, the jobs are not).
## Several runners, and whether they should share a job network
What *is* shared, with one compose file or five, is the job network: every
runner whose config names `dejima-ci` puts its jobs on that bridge. With N
runners at `capacity: 1` you get up to N jobs there at once, able to reach each
other and each other's service containers.
- **One shared job network** — the default, and why several runners can share
one config file. Every job on it has the same egress allowlist and none can
reach the host or the runners, so the added exposure is strictly job-to-job.
If the concern is a prompt-injected session reaching *outward*, this already
covers it.
- **One per runner** — worth it if concurrent jobs come from repositories that
should not interfere, or if service containers hold anything a neighbour
should not see. Costs a config file and a `docker network create` each, and
the firewall rules looped over each bridge.
The `ci-allow` ipset is shared either way — it holds destinations, not
per-bridge state, so one set and one dnsmasq serve every bridge. The wrinkle
is DNS: `daemon.json`'s `dns` is a single daemon-wide address that would sit
on one bridge and be unreachable from the others. Give dnsmasq a stable
address on a dummy interface, point `daemon.json` there, and let each
bridge's `INPUT` rule allow port 53 to it.
## Why the privileged runner is separate
It is handed the Docker daemon and the host's network namespace, because one
job cannot work without them and neither grant can be reduced without breaking
it. A job on it can take the machine it runs on — so that machine should exist
for nothing else. A separate compose file makes a separate host the default,
rather than something to disentangle later.
Where the two tiers do share a machine, be clear about what that is: **label
tiering, not host separation.** A job denied the socket still shares a kernel
with a runner that has it. A real reduction over granting everything to
everything; not isolation between the tiers.
It needs no bind for `/tmp/dejima-ci` — the job container gets that path
because the workflow asks for it and `container.valid_volumes` permits it, and
the host daemon resolves the source host-side.
## Why public egress is not restricted on the node
The Proxmox firewall in [proxmox/README.md](../proxmox/README.md) denies the
privileged VM every RFC1918 destination and leaves the public internet to the
default policy. That asymmetry is deliberate, and it is the opposite of what
an earlier version of that runbook said.
An address allowlist there would break on every CDN rotation. A port allowlist
buys nothing: an earlier version permitted only tcp/80, tcp/443 and udp/123,
which broke apt on a first install, and each breakage presented as a network
fault rather than as a firewall rule. Anyone who owns that box tunnels over
443 like everybody else, and neither restriction would hold against VM root in
any case.
Deny-by-default earns its keep where the protected set is enumerable, which is
the LAN. The internet is not, and pretending otherwise produced a rule set
that looked strict and only got in the way.
Note what that leaves. On the privileged tier there is no egress allowlist in
front of job containers either — they are in the host's network namespace, so
nothing is in front of them. A privileged job reaches the whole public
internet by design. What it does not reach is your LAN.
## Why the resolver is set per-runner, not daemon-wide
`daemon.json` has a `dns` key, and pointing it at the allowlisting resolver is
the obvious way to make job containers use it. It is also wrong here, and the
failure is not obvious.
That setting applies to **every container the daemon starts**. On a host that
also runs the privileged tier, that includes containers created by its jobs
over the Docker socket — which are not job containers and were never meant to
be behind the CI resolver.
A dejima sandbox is exactly such a container, and it has a firewall of its
own. It allowlists precisely one host address: its own bridge gateway, taken
from its default route. A resolver on a *different* bridge is therefore
rejected by the sandbox's own rules, so nothing inside it resolves — and the
symptom is not "DNS is misconfigured" but the sandbox's startup self-check
reporting that a host it allows is unreachable.
`--dns` in `container.options` scopes the resolver to the containers that
runner starts, which is the set that should have it.
## The cache server
`act_runner` enables it by default, and it binds on the host *specifically* so
job containers can reach it — an inbound service exposed to untrusted code. The
firewall here blocks exactly that. Leave it off, or add a rule knowing what you
are exposing.
examples/act-runner-ci.yaml → examples/act-runner-confined.yaml
+4 -4
View File
@@ -1,5 +1,5 @@
# act_runner configuration for a runner that executes agent-authored code and
# is granted nothing to execute it with. This is the ordinary tier: everything
# is granted nothing to execute it with. This is the confined tier: everything
# except the one job that genuinely needs a Docker daemon belongs here.
#
# Written against act_runner v0.6.1. Check yours with `act_runner --version`,
@@ -8,9 +8,9 @@
# one without anything in a job's output looking different.
#
# Not runnable as written: pick your own network name, subnet and label, and
# see ci/README.md for the order things have to be created in.
# see this repository's README for the order things have to be created in.
#
# See docs/ci.md for why each of these is here.
# See dejima's docs/ci.md for why each of these is here.
log:
level: info
@@ -72,7 +72,7 @@ container:
# A network created ahead of time, with a named bridge, so that the
# interface name and gateway are stable enough to write firewall rules
# against. Left empty, act_runner creates one per job and there is nothing
# fixed to match on. See ci/iptables-ci.example.sh.
# fixed to match on. See bin/dejima-ci-firewall.
#
# NOT `host`: that would put jobs in this machine's network namespace, where
# no bridge-based filtering applies to them at all.
examples/act-runner-runtime.yaml → examples/act-runner-privileged.yaml
+3 -3
View File
@@ -14,14 +14,14 @@
# - holds no credentials beyond its own registration token,
# - is rebuilt from a snapshot on a schedule.
#
# Sharing a machine with the ordinary runner (ci/act-runner-ci.example.yaml)
# Sharing a machine with the confined runner (examples/act-runner-confined.yaml)
# gives you label tiering, not host separation: a job denied the socket still
# shares a kernel and a network with a runner that has it. That is a real
# reduction and worth doing if one machine is what you have - just do not read
# it as isolation.
#
# Written against act_runner v0.6.1. Not runnable as written.
# See docs/ci.md.
# See dejima's docs/ci.md.
log:
level: info
@@ -46,7 +46,7 @@ cache:
enabled: false
container:
# Left unset ON PURPOSE, which is the opposite of the ordinary runner. Unset
# Left unset ON PURPOSE, which is the opposite of the confined runner. Unset
# means act_runner mounts /var/run/docker.sock into job containers, and the
# job here builds an image and boots real containers over it.
#
examples/dnsmasq-ci.conf
+3 -3
View File
@@ -1,5 +1,5 @@
# The static half of the job network's resolver. Install as
# /etc/dnsmasq.d/dejima-ci.conf - see "Where each file goes" in ci/README.md.
# /etc/dnsmasq.d/ci-allow.conf - written by install.sh. See this repository's README.
#
# There are no `server=` lines here on purpose. Every allowed name is rendered
# into a SECOND file, /etc/dnsmasq.d/ci-allow.conf, by
@@ -16,7 +16,7 @@
# off encoding data into query names to a domain you never allowed.
#
# Not runnable as written: the addresses below are RFC 5737 documentation
# ranges. Substitute your own, and read the subnet note in ci/README.md before
# ranges. Substitute your own, and read the subnet note in etc/config.example before
# picking one - "private" is load-bearing here, not tidiness.
# bind-dynamic, not bind-interfaces. Both keep this resolver off every other
@@ -52,7 +52,7 @@ ipset=/#/ci-allow
# never on any list.
#
# This is keyed on the private ranges, which is one of the reasons the job
# bridge itself should use one - see ci/README.md.
# bridge itself should use one - see this repository's README.
stop-dns-rebind
# A self-hosted forge legitimately resolves to a private address, which
examples/docker-compose.yml → examples/docker-compose-confined.yml
+11 -11
View File
@@ -1,8 +1,8 @@
# The ordinary runners: several on one host, sharing one config file and
# The confined runners: several on one host, sharing one config file and
# therefore one label, so a job naming that label lands on whichever is free.
#
# The privileged runner is NOT here. It has its own compose file, because it
# belongs on its own host - see ci/docker-compose-runtime.example.yml.
# belongs on its own host - see examples/docker-compose-privileged.yml.
#
# The thing to understand before reading the rest: compose's networks and the
# job network are different worlds. Every service below is an act_runner
@@ -10,14 +10,14 @@
# created by the Docker daemon, on the network named by `container.network` in
# each runner's config.yaml, and compose knows nothing about them. Attaching a
# service here to a network does not put its jobs there, and vice versa. See
# ci/README.md for what that means for isolating jobs from each other.
# docs/reference.md for what that means for isolating jobs from each other.
#
# Not runnable as written: substitute your own instance URL, image tag and
# subnet, and create the job network first (ci/README.md, step 2).
# subnet, and create the job network first - install.sh does it for you.
name: dejima-runners
# Shared by every ordinary runner. Volumes are per-service rather than here,
# Shared by every confined runner. Volumes are per-service rather than here,
# because each runner needs its own /data - and YAML merge keys merge mappings
# but not lists, so a volumes: in a service replaces this one rather than
# adding to it. Repeating them is the honest form.
@@ -44,7 +44,7 @@ x-ci-env: &ci-env
GITEA_RUNNER_REGISTRATION_TOKEN: ${GITEA_RUNNER_REGISTRATION_TOKEN:-}
services:
# --- ordinary runners -----------------------------------------------------
# --- confined runners -----------------------------------------------------
# All three share one config file, and therefore one label and one job
# network. That is what makes them a pool: a job naming dejima-ci lands on
# whichever is free. Only the name and the data volume differ.
@@ -57,11 +57,11 @@ services:
volumes:
# The runner's OWN access to the daemon, so it can create job
# containers. NOT the same setting as container.docker_host, which
# decides whether job containers get the socket - see ci/README.md.
# decides whether job containers get the socket - see docs/reference.md.
- /var/run/docker.sock:/var/run/docker.sock
# Read-only: a runner has no reason to rewrite the file deciding what
# its jobs are granted.
- ./act-runner-ci.yaml:/config.yaml:ro
- ./act-runner-confined.yaml:/config.yaml:ro
# Its registration identity. One per runner, never shared - two runners
# cannot share a .runner.
- ci-1-data:/data
@@ -74,7 +74,7 @@ services:
GITEA_RUNNER_NAME: dejima-ci-2
volumes:
- /var/run/docker.sock:/var/run/docker.sock
- ./act-runner-ci.yaml:/config.yaml:ro
- ./act-runner-confined.yaml:/config.yaml:ro
- ci-2-data:/data
ci-3:
@@ -85,7 +85,7 @@ services:
GITEA_RUNNER_NAME: dejima-ci-3
volumes:
- /var/run/docker.sock:/var/run/docker.sock
- ./act-runner-ci.yaml:/config.yaml:ro
- ./act-runner-confined.yaml:/config.yaml:ro
- ci-3-data:/data
networks:
@@ -96,7 +96,7 @@ networks:
# would protect nothing.
#
# The job network is deliberately absent. It is created by hand with a fixed
# bridge name so the firewall can match on it (ci/README.md, step 2), and
# bridge name so the firewall can match on it (install.sh does this), and
# act_runner attaches job containers to it by name. Declaring it here would
# invite attaching a service to it, which is exactly what must not happen: a
# process holding the Docker socket has no business sharing a network with
examples/docker-compose-runtime.yml → examples/docker-compose-privileged.yml
+4 -4
View File
@@ -9,7 +9,7 @@
# nothing else: no other service, no route to a LAN or management network, no
# credentials beyond its own registration token, rebuilt from a snapshot on a
# schedule. That is why this is a separate file rather than another service
# alongside the ordinary runners - a separate host is the intended
# alongside the confined runners - a separate host is the intended
# arrangement, and a compose file that assumes it is harder to drift away from
# than one that has to be split later.
#
@@ -18,7 +18,7 @@
# with a runner that has it. That is a real reduction over granting everything
# to everything, and it is not isolation between the tiers.
#
# Pair with ci/act-runner-runtime.example.yaml. Not runnable as written:
# Pair with examples/act-runner-privileged.yaml. Not runnable as written:
# substitute your own instance URL and image tag.
name: dejima-runtime
@@ -52,10 +52,10 @@ services:
# The runner's own access to the daemon, so it can create job
# containers. On this tier the config file ALSO passes the socket into
# those containers, which is the grant that makes this host sacrificial;
# on the ordinary tier it does not. Same bind, different consequence,
# on the confined tier it does not. Same bind, different consequence,
# decided entirely by container.docker_host in the config.
- /var/run/docker.sock:/var/run/docker.sock
- ./act-runner-runtime.yaml:/config.yaml:ro
- ./act-runner-privileged.yaml:/config.yaml:ro
# Its own registration identity. Never shared with another runner.
- runtime-data:/data
# No bind for /tmp/dejima-ci. The job container gets that path because
install.sh
+4 -1
View File
@@ -1,4 +1,7 @@
#!/usr/bin/env bash
# SPDX-License-Identifier: GPL-3.0-or-later
# Copyright (C) 2026 Brad Anderson
#
# Install dejima-ci-allowlist and walk through its configuration.
#
# Idempotent: re-running and pressing Enter at every prompt changes nothing.
@@ -135,7 +138,7 @@ say "=============================================================="
say
say "A Docker network for job containers to run on, which this installer will"
say "create. Name it whatever you like - it only has to match container.network"
say "in each ordinary runner's config.yaml."
say "in each confined runner's config.yaml."
say
say "A runner configured for host networking (container.network: host) needs"
say "no job network - its jobs are in the host's own namespace, so no bridge"
proxmox/README.md
+11 -10
View File
@@ -28,9 +28,10 @@ root-equivalent on their host. So the working assumption is not *if* but
*when*: a compromised job has root on this VM. Everything below follows from
taking that seriously rather than hoping.
Two VMs if you are following [the tiering](../README.md#register-scope-is-the-boundary-the-label-is-not):
one for the privileged runner, one for the ordinary ones. The steps are the
same; only the runner you register on each differs.
Two VMs if you are following [the
tiering](../README.md#two-tiers-and-you-must-pick-one): one for the privileged
runner, one for the confined ones. The steps below are the same for both; what
differs is the `--role` in §4 and the runner you register in §5.
**The firewall goes on the node, not in the guest.** Root in the VM can flush
any rule set it can see. It cannot reach one applied to its own tap device.
@@ -265,7 +266,7 @@ ssh <you>@<vm-ip>
sudo apt-get update && sudo apt-get install -y git # if the install left it out
git clone https://gitbox.branderson.io/branderson/dejima-ci-hardener.git
cd dejima-ci-hardener
sudo ./bin/dejima-ci-bootstrap --role runtime --mgmt <your-management-ip>
sudo ./bin/dejima-ci-bootstrap --role privileged --mgmt <your-management-ip>
```
Two ways this stalls, both from choices made during the install:
@@ -277,13 +278,13 @@ Two ways this stalls, both from choices made during the install:
during the install, so Debian did not add you. See "Leave the root password
blank" above for the fix, or just run these as root from the console.
`--role ci` for the ordinary tier. Either way the script installs packages,
`--role confined` for the other tier. Either way the script installs packages,
hardens sshd, removes listeners a runner has no use for, and sets up
unattended security upgrades. On `--role ci` it then runs `install.sh` for the
job allowlist, its resolver and its firewall; on `--role runtime` it
deliberately does not, because those work by sitting in front of job traffic
on a bridge and this tier's jobs are in the host's network namespace, where
there is no bridge to sit on.
unattended security upgrades. On `--role confined` it then runs `install.sh`
for the job allowlist, its resolver and its firewall; on `--role privileged`
it deliberately does not, because those work by sitting in front of job
traffic on a bridge and this tier's jobs are in the host's network namespace,
where there is no bridge to sit on.
It does not register the runner — that needs a token from your forge — and it
does not enable the host's own egress policy until you have filled in
test/allowlist.bats
+2
View File
@@ -1,4 +1,6 @@
#!/usr/bin/env bats
# SPDX-License-Identifier: GPL-3.0-or-later
# Copyright (C) 2026 Brad Anderson
#
# Drives bin/dejima-ci-allowlist against a stubbed curl and dnsmasq, so the
# properties being checked are the ones that decide whether this program is