.gitbay/wiki/Threat-Model.org

  1#+title: gitbay threat model
  2
  3What the forge trusts, what it refuses to do, and where the boundaries
  4are. This is the reference for security review; it complements the audit
  5log and hardening notes in [[Admin]].
  6
  7* What gitbay never does
  8
  9- *Execute repository content.* Git object contents are never run. Hooks
 10  are gitbay's own binary, invoked by git; they compute facts and ask the
 11  daemon over a unix socket. Repo files are only ever read.
 12- *Hold a signing key.* There is no server-side signing key. "Verified"
 13  means a signature made by a key the *user* registered — the server
 14  never vouches for a commit it did not receive already signed. Merge
 15  commits the server creates are honestly =unsigned=.
 16- *Serve repository HTML on its own origin as active content.* Raw file
 17  serving is =text/plain= with =nosniff=. Rendered markdown/org is
 18  sanitized (bluemonday) and served under a CSP that forbids scripts.
 19- *Put secrets in argv, URLs, or logs.* Import and mirror credentials,
 20  registration invites, and API tokens travel on stdin or in request
 21  bodies, never as command arguments (visible in =/proc=) or query
 22  strings. Tokens are stored only as SHA-256 hashes.
 23- *Name a mail recipient in the log.* A queued mail is logged by its
 24  queue row id, never by address, and the relay's own error is redacted
 25  before it is logged because a rejection usually quotes the address it
 26  rejected. The unredacted error and the address stay on the row, which
 27  an instance admin reads on =/admin=: the database holds who, the log
 28  holds which. One rule for every mail type — notifications, dependency
 29  reports and login links all drain the same queue (krz/gitbay#173).
 30- *Confirm the existence of private repositories.* Every surface answers
 31  "not found" identically for a private repo and a nonexistent one — web
 32  pages, git transport, control commands, release asset downloads.
 33
 34* Trust boundaries
 35
 36- *SSH public key = identity.* The SSH username is ignored; the presented
 37  key's fingerprint resolves to an account. Key uniqueness is global.
 38- *Per-instance trust.* Email verification and key registration are local
 39  to an instance and never transfer. Account migration re-registers keys
 40  and re-verifies emails on the target by design.
 41- *The control plane is one authenticated channel* (SSH), fully usable
 42  from stock OpenSSH. The JSON API fronts the same command registry with
 43  bearer tokens minted only over SSH; git transport never runs over it.
 44- *Anonymous surfaces* — HTTPS clone of public repos, =git://= where
 45  enabled, the read-only web UI — carry no credentials and expose only
 46  public data. HTTP push is refused via a pkt-line =ERR=, never a 401.
 47
 48* Attacker-controlled parsers
 49
 50Every parser that eats bytes from a pusher, a key registrant, or an
 51anonymous client has a fuzz target and must never panic:
 52
 53- =internal/protocol= — the SSH command tokenizer (fuzzed against argv
 54  round-tripping).
 55- =internal/gitd= — the =git://= pkt-line reader.
 56- =internal/sig= — the commit parser, the SSHSIG armor decoder and blob
 57  parser, and the OpenPGP armored-key reader.
 58
 59Run =deploy/audit.sh= to exercise them plus =go vet= and =govulncheck=.
 60
 61* Secret handling
 62
 63Tokens (web sessions, login links, email verification, API bearer tokens,
 64deploy/CI) are random 256-bit values. Only their SHA-256 hash is stored,
 65and verification is a database index lookup on that hash — the secret
 66itself is never compared in Go, so there is no timing oracle to exploit.
 67Webhook payloads are signed outbound with HMAC-SHA256; the forge verifies
 68no inbound HMAC.
 69
 70* Network-facing request forgery
 71
 72Anything that makes the *server* open an outbound connection to a
 73user-supplied address — webhook delivery, GitHub-history import
 74=--api-base=, mirror remotes — passes the same SSRF guard: the scheme
 75must be http/https and, unless =webhooks.allow_local= is set, the
 76resolved address must not be loopback, private, or link-local. The
 77webhook dialer re-checks at connect time so a DNS answer that changes
 78after validation still cannot reach private space. Redirects are never
 79followed.
 80
 81* Rendering pushed markup
 82
 83Rendered markup is attacker-controlled: a README, a wiki page and a
 84profile's about text are all whatever someone pushed or typed. The risk
 85is not only what the output contains but what the *parser* is willing to
 86go and fetch — the filesystem counterpart of the SSRF guard above.
 87
 88Org is rendered by go-org, whose default configuration resolves
 89=#+INCLUDE:= and =#+SETUPFILE:= targets with =os.ReadFile=. Both are
 90refused outright (=orgConfig()= in =internal/httpd=): the file is never
 91opened and the keyword stays the inert text it already was, so the rest
 92of the document renders normally. There is no safe subset to allow
 93instead — an absolute path skips go-org's relative-path join, a relative
 94one resolves against the daemon's working directory, and the content
 95came from a git object rather than a checkout, so there is no directory
 96to scope a read to. Markdown is goldmark, which has no include
 97mechanism. go-org's parse warnings are discarded rather than logged, so
 98pushed content cannot write to the server's log.
 99
100Org output is then sanitized (bluemonday UGC policy) because go-org
101passes raw HTML through — export blocks and inline export snippets —
102while goldmark drops it and needs no pass. The policy admits chroma's
103short token classes and nothing else.
104
105* Web responses
106
107Every response carries =Content-Security-Policy= (no scripts, no plugins,
108no embedding; inline styles allowed for chroma and label chips; images
109from any origin so external README images render), =X-Frame-Options:
110DENY=, =X-Content-Type-Options: nosniff=, =Referrer-Policy: no-referrer=,
111and =Strict-Transport-Security= when TLS is on. The UI needs no
112JavaScript, so =script-src 'none'= costs nothing.
113
114* The CI runner
115
116=gitbay-runner= is the one component that executes repository content.
117gitbayd never does: it reads =.gitbay/ci.yml= and queues a build, and a
118runner, polling over SSH, clones the commit and runs its steps.
119
120- *What the runner holds.* A key of scope =runner=, which the dispatcher
121  confines to =runner next=, =runner log= and =runner done= and to
122  read-only git, and which claims, logs and finishes builds only for
123  the repositories it is attached to (=repo runner add=). A step that
124  reads the key off the disk gets exactly that: it cannot administer
125  the instance, push, read a repository the runner's account cannot, or
126  touch another repository's builds. An admin key still works for the
127  runner protocol so an operator can rotate at their own pace; a runner
128  host should not hold one. Untrusted builds are skipped unless the
129  runner asks with =-untrusted=, so a runner on a user's machine never
130  executes a stranger's branch by default.
131- *What a build sees.* The commit, the =GITBAY_*= variables and the
132  repository's secrets — unless the head came from another repository.
133  A merge request from a fork is built in the target as untrusted, with
134  no secrets, so a stranger's branch cannot read the target's deploy
135  credentials. The step environment is *constructed*, not inherited: a
136  build gets =PATH=, =HOME=, =LANG=, =CI=, its own variables and its
137  secrets, and nothing the operator set on the service. =HOME= is a build
138  home under the runner's =-workdir=, not the runner's own home, so a
139  build cannot read the =.netrc=, =.npmrc= or =.gitconfig= where tools
140  keep credentials. That home is shared by every build on the runner —
141  one build can poison a cache another reads, which is no more than
142  anything a step can already do as this user, and is what isolation
143  (krz/gitbay#144) is for.
144- *Where it runs.* Steps run in a rootless podman container, one per
145  job, with the workspace bind mounted and nothing else. The clone
146  happens outside it with the runner's key, so the container never sees
147  =GIT_SSH_COMMAND=, the key, or the runner's environment. =-isolation
148  none= runs steps on the host as before, for an instance where every
149  repository is trusted; there is no automatic fallback to it — a runner
150  configured for podman that cannot find one refuses to start, because
151  dropping isolation silently is worse than a stopped runner. The
152  systemd drop-in still adds =ProtectSystem=full= and the cgroup
153  protections, and =-repos= still limits a runner to named
154  repositories. =ProtectKernelTunables= is *off*: it overmounts =/proc=
155  in the unit's namespace and the kernel then refuses a proc mount in
156  any child user namespace, which every rootless container needs; the
157  container masks the same paths for the build itself.
158  =NoNewPrivileges= is *off*: rootless podman sets up its
159  namespace with the setuid =newuidmap=, which that flag blocks, so the
160  choice is between it and containers at all. Containers are the stronger
161  boundary — the flag constrained a process that was already running
162  arbitrary repository code, and under podman that code no longer runs in
163  the runner's process context. Under =-isolation none= there is no
164  container and the flag should be on.
165- *Images are provisioned by the operator, not fetched by a build.* The
166  runner passes =--pull=never=, so =image:= chooses among what the host
167  already has rather than naming anything on the internet. On an
168  instance with open registration that is the difference between a
169  curated set and arbitrary code from a registry nobody vetted.
170
171Under =-isolation none=, anything a step can do as the runner's user a
172pushed =ci.yml= can do. Under podman a step is confined to its
173container, the bind-mounted workspace and the repository's own build
174home, so what a build leaves in a cache is read only by later builds of
175the same repository. Treat the runner host as executing untrusted code
176all the same: keep it off the daemon's host where the database lives,
177or scope it to repositories whose writers you trust. gitbay.org does
178the latter — its runner builds only the repositories the operator
179names.
180
181* What has not been audited
182
183The 2026-09 sweep (krz/gitbay#149) read this repository against the
184claims above. Its two findings — review verdicts from accounts without
185write access deciding merge gates (#147), and control commands having no
186write rate limit while the API had one (#148) — are fixed, as is #173,
187found afterwards in the same milestone. What it did not reach, and why,
188is recorded here rather than in a closed issue:
189
190- *The host.* systemd sandboxing, sshd on 2222, the firewall,
191  unattended-upgrades, fail2ban, disk and file modes, restic append-only
192  credentials. Reading production configuration is a separate exercise
193  from auditing the source, and needs doing against bay1.
194- *The supply chain beyond =govulncheck=.* No review of what the
195  dependency set is, who maintains it, or what a compromised release of
196  any of it would reach.
197- *The runner host as an execution environment.* Nobody has tried to
198  escape what is there. #144 covers the missing isolation.
199- *Timing and traffic analysis.* Token comparison is a hash index lookup
200  by design, but nothing has been measured.
201- *Denial of service by resource exhaustion* beyond rate: large pushes,
202  pathological diffs, deep histories, zip bombs in LFS.
203
204A sweep is a point in time. This section says what a reader should not
205assume has been checked.
206
207* Residual risks, accepted
208
209- External images in rendered READMEs and profile about text load from
210  their origin (no image proxy), which the author can use as a tracking
211  pixel against a viewer. A profile is the wider surface of the two: it
212  is linked from every commit and issue its owner touches. Documented;
213  proxying is future work.
214- Backups are consistent per the DB-snapshot-first ordering but are not a
215  single atomic snapshot; a few orphaned git objects are possible and
216  harmless (see [[Admin]]).
217- A global signature-verification epoch over-invalidates the cache on any
218  trust-input change. Correct, not a leak; a performance tradeoff.
219- A build's secrets are environment variables inside its container, so
220  they are visible to =podman inspect= as the runner's user — the same
221  user that already holds them in memory. They reach podman through a
222  0600 env file rather than argv, since =/proc= is world-readable.