Files
DodoSSH/docs/design-import-gaps.md
T
jaap-jan 95816de0c5 Share a vault with a team, without the server holding a key
M3's teams, sharing and ACLs. Teams with roles, a public-key directory, the
append-only key log served for clients to check it against, team-owned vaults,
and vault key grants wrapped by a client and stored opaquely by the server.
VaultAccessService resolves team membership to PermissionFlags, so a viewer may
pull and may not push; the desktop client reads and syncs every vault it holds
a key for, and a real TEAMS screen replaces the one that said it did not exist.
No migration: team, team_membership, vault.team_id and vault_key_grant have all
been there since the first one, which is what carrying two unused tables bought.

Membership is authorisation. A grant is access. The obvious model is one
concept — "access", with a role attached, handed out by the server — and this
architecture cannot implement it: a vault key is sealed to each member's X25519
key, and only a client holding the plaintext can seal it for somebody else. So
"give Bob access" decomposes into a database write and a wrap, which happen on
different machines. Adding a member makes the server serve them the vault; it
cannot make it readable. VaultSummary.WrappedVaultKey is null in the meantime
and the vault appears in their list saying it is waiting for a key, because
hiding it until a grant existed would have been tidier and would have implied
the server was the thing granting access. The screen says the same thing after
every add, in the status line. ADR 0009 records the whole decision.

Sharing verifies or refuses. A directory lookup is a claim by the server about
a third party's public key, and wrapping to an unverified claim hands the vault
to whoever made it — no amount of transport security helps, because the server
is inside the threat model. KeyLogAudit reads the whole log, recomputes every
entry's hash from its own contents, checks the chain from genesis, and refuses
unless the offered key appears in it unchanged. There is no override flag: one
that exists gets used on the day the log is briefly unreachable, and the
resulting grant is indistinguishable from a correct one afterwards. What it
still cannot promise is that the key is the right person's, so the fingerprint
comes back for an out-of-band comparison and the success message says so every
time. A test corrupts the fake server's log by one byte and watches the client
refuse rather than warn.

The roles are only the ones that are enforceable. There is no ConnectOnly,
despite the design asking for one and TeamRole having room: SSH terminates on
the client, so a session needs the credential's plaintext on that machine, and
"may connect but may not read the key" cannot be enforced here. Shipping it as
an option in a dropdown would have been a lie. Connect rides along with Read
and is documented as an interface hint. Removal is named for what it does — it
revokes grants and flags the vault for rekey, and claims nothing about what is
already on somebody's laptop.

Three things are deliberately absent, and each is a refusal rather than an
omission. The rekey itself, because re-wrapping every item's data key under a
new vault key needs a client holding the current one; the server records that a
rotation is owed and the interface reports it, which is more honest than a
button that only appears to do it. Ownership transfer, because allowing an
owner to be removed without one leaves a team nobody can administer. And
cross-vault host key trust: a pin in a team vault is listed but not consulted
at connect time, because any member with Write could otherwise pre-approve a
fingerprint another member's client then trusts silently for a host in their
own vault. Scoping trust properly needs a scope on the SSH connect path, which
IKnownHostStore has not got; until then the narrow direction is the safe one
and the cost is in the README rather than hidden.

Reading now spans vaults and writing still does not. Every list on the vault
and hosts screens covers each vault the keyring opened, rows carry the vault
they came from, and an edit goes back to that vault rather than to the active
one — writing it to the active vault would fork the item and only show up when
a colleague wondered why their change never arrived. A new item goes wherever a
picker says, defaulting to the personal vault and never moving on its own,
because an item filed into a team's vault is visible to that team and moving it
back means deleting and retyping. The sidebar heading stops naming one vault
once there are two, and each row names its own.

The server checks what it can and nothing it cannot. It will not record a grant
for a key its recipient no longer holds, for a superseded generation, or for
somebody who is not in the team — each of those would otherwise surface days
later at the far end as a tag failure indistinguishable from corruption. It
does not verify the wrap or the signature, and the grant service says so: that
would be a convenience and never the boundary, and would put an asymmetric
implementation on a machine that is supposed to hold no keys.

Two bugs the tests found. TeamsViewModel's busy gate blocked its own reload, so
a team created a moment earlier was missing from the list it had just been
added to. And syncing every vault turned a failure from an exception into a
report, which made a background pass announce an unreachable vault once a
minute — the exact behaviour AnAutomaticPassThatFails_LeavesTheStatusAlone
exists to prevent. The fact is recorded and the message swallowed, as it was
before; pressing Sync still names the vault and the reason.

Also fixes a build break this branch started with: QuickConnectTests was never
updated when M2 added ISftpSessionFactory to the shell's constructor, so
nothing built at all.
2026-07-31 12:18:28 +02:00

25 KiB
Raw Blame History

What the design asked for and this build has not got

The Avalonia client's interface was rebuilt from a Claude Design project — a five-screen spec covering the chrome, hosts and terminals, file transfer, the vault, teams, and preferences.

Most of it landed. This file is the rest: every element of that design with nothing behind it, which project each piece would have to land in, and what the shipped interface does instead. That last column is the important one. This codebase states its limits rather than implying them — the README does it for revocation, for relay session recording and for what locking does not do — so nothing here was rendered with invented data to make a screenshot look finished. Where a feature does not exist, the interface either omits it or says so.

Read alongside the milestone plan. Most of what follows is not a defect; it is M2 and M3 arriving in a design before it arrives in the code.

The shape of the problem

Three facts explain nearly every row below.

An SSH connection here opens exactly one channel. ISshConnection offers OpenShellAsync and nothing else (src/DodoSSH.Client.Ssh/SshConnection.cs). No port forwarding, no ProxyJump. That one fact removes the FORWARDS chip, the status bar's port list, and every via bastion-eu in the design.

It used to remove the file-transfer screen too. M2 did not lift the restriction — it worked around it: file transfer is a separate connection rather than a second channel, because SSH.NET's SftpClient owns its own transport. See File transfer, which is the one place in this document where what shipped differs from what the row predicted.

Teams are schema and nothing else. Built in M3. The team and team_membership tables were there from the first migration with nothing reading them, and VaultAccessService.ResolveAsync denied every vault that was not the caller's own. Both changed in M3 and neither needed a migration, which is what carrying two unused tables bought. See Teams. What has not changed is the split underneath: the server decides what it will serve, and only a client can decide who can read it — so "shared with" is two facts on this screen, not one.

The client has no preferences store. It writes exactly two files — cache.db and device.key — and the cache has six tables, none of them settings. Nothing on the design's TERMINAL preferences panel can be saved, and there is no frame on the terminal data plane that would carry a change to the renderer anyway.

One thing cuts the other way and is worth knowing before planning any of this: the wire protocol already reserves the slots. SyncEntityType (src/DodoSSH.Contracts/SyncEntityType.cs) has HostGroup = 4, Tag = 5, HostTag = 6, HostCredential = 7, Snippet = 8 and PortForward = 9 — reserved, unused, and already covered by the AAD resource-type table. Groups, tags and snippets are new item types on an existing protocol rather than a protocol change.


Chrome — titlebar, nav rail, status bar

Design element Layer What it would take What ships instead
Org chip dodotech / platform contracts + server An organisation above teams, which does not exist — VaultSummary.TeamId is no longer always null since M3, but a team is not an org and there is exactly one tenant per deployment. The vault's own name, and the account this machine is enrolled as. Team names are on the TEAMS screen, where they are about something.
SYNCED dot, always green client-app Nothing — the design's claim is simply unconditional. Green only when a connection is held, the last sync pass actually reached the server, and the outbox is empty; otherwise UNREACHABLE, the count of changes still waiting, or OFFLINE. Holding an IVaultServer proves a sign-in once succeeded and nothing more, so a laptop whose lid has been shut all afternoon still has one — reachability comes from the outcome of the last pass. A permanently green light is the same as no light.
VAULT SYNCED 11:02 client-session StoredSyncState.LastPulledAt/LastPushedAt are persisted, but VaultSession exposes the store as internal. A property away. Omitted. The one honest sync fact — the outbox depth — is in the titlebar and the status bar.
⌘K command palette running commands client-domain A snippet or saved-command item type (SyncEntityType.Snippet = 8 is reserved). Ctrl+K opens a real host search that connects on Enter. The box says "search hosts", not "search hosts · run command".
Status bar · via bastion-eu client-ssh Jump-host execution. See below. Omitted.
Status bar port forwards client-ssh Port forwarding. See below. Omitted.
Status bar sftp · 2 transfers client-app Nothing now — file transfer is built. What is missing is the count reaching the status bar, which is a screen away from where the queue lives. Omitted from the status bar. The queue itself is on the FILES screen, with a row per transfer.
Status bar locks in 09:41 client-app An idle auto-lock. See preferences below. Omitted.
IBM Plex Mono / IBM Plex Sans ui Shipping the font files as AvaloniaResource and registering them. The design loads them from Google Fonts, which a desktop app cannot. Inter (already embedded) for prose, and the system monospace stack the terminal already names. Named once in App.axaml as MonoFont, so the substitution is reversible in one place.
⌘K, ⌥↵ ui Nothing; the design is Mac-flavoured. CTRL K. Development is Windows-only today (docs/platform-flags.md).

Also worth knowing: custom chrome went in as SystemDecorations="BorderOnly", not by extending the client area. ExtendClientAreaChromeHints was removed in Avalonia 12, so extending leaves the system caption buttons and window title drawn on top of the application's own — two sets of window controls, and "DodoSSH" printed twice. Measured, not assumed.


Hosts + terminal

Design element Layer What it would take What ships instead
Tag chips (nginx, eu, pg16) client-domain A tag item type and a host-tag join. Both reserved on the wire (Tag = 5, HostTag = 6), neither implemented, plus a payload schema bump on HostSecret. Omitted. The filter box searches name, address and notes instead.
Groups PRODUCTION / STAGING / PERSONAL client-domain A host-group item type (HostGroup = 4, reserved) or a group field on HostSecret. One heading per vault — the only grouping a host actually has. Since M3 there is genuinely more than one when somebody is in a team, which is what that row always predicted.
Group badge TEAM·PLATFORM server Built in M3. The vault's name on each row, and the personal vault ordered first. Not the team's name: two of a team's vaults would then carry the same badge and the badge would be naming the wrong thing.
Per-host status dot, three colours client-ssh The amber state would mean "reachable but not connected", and nothing here ever probes a host. Two states, both real: green when a terminal is open on that host, grey when not.
· ⤷ bastion-eu in the host subtitle client-ssh Jump hosts are data-only. HostSecret.JumpHostIds is a JumpChain that is stored, encrypted, synced and three-way merged — and nothing reads it at connect time. SshConnectionRequest carries one host. Omitted. The stored chain is preserved untouched by every edit.
SPLIT ⌘D and side-by-side panes client-ssh + ui The renderer stacks panes and shows one (terminal.css: .pane { position:absolute; inset:0; display:none }). Tiling needs a real pane geometry and a splitter. Omitted. Tabs ship instead, over the same one-WebView multiplexing.
⇄ FORWARDS · 2 client-ssh Port forwarding. SyncEntityType.PortForward = 9 is reserved; nothing in the SSH layer forwards anything. Omitted.
SNIPPETS panel, to run client-domain A snippet item type (Snippet = 8, reserved). Omitted.
Broadcast to all panes (⌥↵) client-ssh Input is routed strictly by session id in TerminalDataPlane.Dispatch; there is no fan-out. Needs splits first. Omitted.
Pane header 24ms client-ssh Round-trip measurement. SSH.NET offers no RTT API. Omitted.
Pane header aes256-gcm client-ssh The closest miss on this list. SshNetConnection holds the SshClient, so ConnectionInfo.CurrentServerEncryption is right there — it just is not on ISshConnection or surfaced by TerminalWorkspace. Omitted; the tab strip shows the account and endpoint actually dialled.
Pane header showing the running command and following client-ssh The host moves opaque bytes and never parses terminal output. Would need shell integration (OSC 133) on the remote. Omitted.
A local · zsh tab client-ssh Every session here is an SSH channel. Needs ConPTY and a second session kind. Omitted.
Tab strip + button ui Not missing so much as redundant: the real operation is select a host, press Connect, which the sidebar already is. Omitted. Connect opens a tab; Ctrl+K opens one by name.
Terminal font size (--termfs, 1116px) client-storage See preferences. Fixed at the renderer's 13px.

File transfer (the design's SFTP screen)

Built in M2. The screen ships: two directory panes, a breadcrumb trail on each, and a queue that moves one file at a time with progress, throughput and resume. What follows is what it does not do, and one thing this document got wrong before it was built.

The correction. The row below used to say an SFTP subsystem channel on ISshConnection was what it would take. SSH.NET does not offer that: SftpClient derives from BaseClient and owns its own transport, and there is no supported way to hand it a session an SshClient already has. So the transfers screen opens a second authenticated connection to the host rather than a second channel on the terminal's. That is visible to a user — the host records a second login, and a host whose password is typed each time asks for it again on this screen — so it is named for what it is: ISftpSessionFactory.OpenSftpAsync is a connect, and it goes through the same host key gate, the same pin and the same two refusals a shell does.

Design element Layer What ships
SFTP itself client-ssh ISftpSession over SSH.NET's SftpClient: listing, stat, offset-based read and write, mkdir, delete and rename. Delete is deliberately not recursive.
Remote listing with NAME/SIZE/MODIFIED/PERMS client-ssh All four. PosixMode renders drwxr-xr-x from the bits SFTP hands over; setuid, setgid and sticky are not shown, because SSH.NET does not surface them and rwx where rws is true would be worse than nothing.
Local listing client-transfer LocalDirectory, which is where this client's System.IO now lives. PERMS is blank on the local side rather than filled with a plausible-looking POSIX mode that is not a fact about a file on Windows.
Transfer queue, progress, throughput client-transfer FileTransferQueue. One transfer at a time, so the rate on a row is the rate of the link rather than a share of it. Throughput is measured over a half-second window, not averaged since the start.
resume supported client-transfer Within a run of the application. Every transfer writes to a .dodossh-part file beside its destination and is renamed into place at the end, so an interrupted one can never be mistaken for a finished one, and RESUME carries on from the part file's own length. A part file found at startup is not resumed: nothing records what wrote it, and resuming on the strength of a name matching is how a corrupt artefact gets delivered with nothing reporting a failure. Making it survive a restart needs the preferences store this client has not got — see below.
Design element Layer What it would take What ships instead
Per-host last directory client-storage Somewhere to persist two panes' navigation state. There is still no settings table. The remote pane opens on the account's home directory, which the server canonicalises during the handshake; the local pane opens on the user profile.
sftp over bastion-eu client-ssh Jump hosts, as above. HostSecret.JumpHostIds is still stored, synced, merged and read by nothing. Omitted.
Overwriting a file that is already there ui A prompt, which means a modal this window has no idiom for. Refused, with the name that is in the way. The remote pane has DELETE and MKDIR so the refusal is not a dead end.
Dragging between the panes ui Drag-and-drop between two ListBoxes, plus a drop target that is a directory rather than a row. Two arrow buttons between the panes, pointing at the pane the file is going to.
Transferring a directory client-transfer Recursive enumeration on both sides, and a policy for what a partial directory means. One file at a time. A directory cannot be selected as a transfer source.

Not to be reached for: CreditWindow is a 256 KiB flow-control window for terminal output, not a transfer primitive. The queue does its own 64 KiB copy loop and shares nothing with the terminal data plane.


Vault

The screen ships and is real: four categories over the vault's four item types, one table, a detail pane, and both editors. What follows is what the design drew around them.

Design element Layer What it would take What ships instead
ACCESS column and your access → CONNECT-ONLY server Per-item ACLs. EncryptedPayload.DataKeyId is documented as the seam for per-item grants in M5. VaultSummary does carry an opaque int Permissions, but nothing gives the bits a meaning on the wire — PermissionFlags itself lives in DodoSSH.Domain — and those are per vault, never per item. There is no ConnectOnly role in TeamRole at all. Omitted. The column shows sync state instead — whether a change is still sitting in this machine's outbox — which the design had no column for.
SHARED WITH · 6 avatars server Built in M3, minus the avatars — no picture is stored anywhere. GET /api/v1/vaults/{id}/grants lists who holds a key. The list lives on the TEAMS screen, beside the members it is about, rather than as a count on an item row: a grant is per vault, and putting a number on an item would imply per-item sharing, which is M5.
"Private key never leaves the vault. Sessions sign through the team agent (dodod)" client-ssh Both sentences are false here, and the second cannot be made true by this architecture. There is no agent of any kind, and the connect path decrypts the private key and hands the bytes to SSH.NET. Omitted. The key editor already says what is true: the key and its passphrase are encrypted here and never reach the server in a readable form.
Scope rail: PERSONAL / TEAM · PLATFORM / TEAM · DATA server Built in M3. VaultAccessService.ListAsync resolves team membership, so a session can hold several vaults. Not a rail, because it would be a selector with nothing to select: every list on the screen already spans every vault this session can read. What replaces it is a picker for where a new item is filed, which is the only vault question with an answer.
SCOPE column server Built in M3. Scope is a property of a vault, never of an item, and that has not changed. Each row names the vault it is in, and rows are grouped by vault.
LAST column (11:02, 1d ago) contracts No last-used timestamp at any layer. VaultItem<TSecret> is (id, secret, version, three sync flags). Omitted.
FINGERPRINT for SSH keys client-domain SshKeySecret has no fingerprint field, and computing one means parsing key formats the type deliberately stores verbatim. The DETAIL column carries what is known — whether a passphrase and a public half are stored. Pins show their real fingerprint, in full and untruncated.
•••• rotated jul 14 for passwords client-domain CredentialSecret has no rotation date or password age. The account the password is for.
IDENTITIES category contracts A new item type, not even reserved in SyncEntityType. Omitted.
CERTIFICATES category, expires 2026-09-01 contracts Same ladder, plus certificate parsing. Omitted.
18 items · 12 shared server The sharing stack. N items · N not synced, counted over the rows on screen.
added by anna@dodotech.dev contracts The server records CreatedByUserId, but SyncChange carries no actor field and no other user's name is fetchable. Omitted, and the detail pane says in one line that items record no author, no timestamps and no sharing.
created 2026-03-14 client-sync Recoverable in principle — entity ids are UUIDv7 and carry a timestamp — but nothing surfaces it. Omitted.
TEST CONNECT ui Connecting is host-scoped, not credential-scoped: there is nothing to test a credential against without a host. Omitted.
REVOKE server Built in M3, as WITHDRAW KEY on the TEAMS screen — because there are now grants to revoke, and a grant is what it acts on. ADR 0001 constrains how it is presented, and it is: the message says it blocks future reads only, and that what they already hold is unaffected. On this screen, still Delete, named for what it does. Deleting an item and withdrawing somebody's key remain different acts.
SSH KEY · ED25519 client-domain No algorithm field, and deriving it means parsing the armour. The type without the algorithm.
One + ADD CREDENTIAL button ui Two buttons, one per kind that can be added. "Credential" is a specific item type in this codebase (a username and a password), so using it as an umbrella word would collide with the vocabulary.
HOST KEYS is the reverse case: a fully-backed, shipped category the design had no slot for. It is in the rail.

Teams

Built in M3. The screen ships: a team list, a members table, the team's vaults, and the two buttons the whole design was really about — add a member, and share a vault key. What follows is what it still does not do, and one thing this document got wrong before it was built.

The correction. The rows below used to describe a screen with nothing behind it, on the grounds that VaultAccessService.ResolveAsync denied every vault that was not the caller's own. That is now the one place that changed, exactly as its remark predicted, and no migration was needed: team, team_membership, vault.team_id and vault_key_grant have all been there since the first migration. What the row did not anticipate is that the interesting half is not the endpoints at all. It is that membership and readability are different things, and the screen is arranged around saying so.

Design element Layer What ships
The team itself server POST/GET /api/v1/teams, plus members, roles and team vaults. Ids are client-chosen, so a create whose response was lost is safe to repeat.
Shared vaults server + client A team owns vaults; each is created with the creator's own grant, because a vault with no grant is a container nobody can open.
Roles contracts + server TeamMemberRole on the wire, numerically pinned to DodoSSH.Domain.TeamRole by a test. Viewer reads, Member writes, Admin and Owner also share and administer.
Members table server TeamMemberSummary, and a directory that resolves an exact email to a public key.
Sharing an item client VaultSession.ShareVaultAsync: verify the recipient's key against the key log, wrap, sign, record. The server stores the wrap and the signature and can check neither.
Design element Layer What it would take What ships instead
CONNECT-ONLY role Nothing that would be true. Connect is a user-interface hint, not a boundary: SSH terminates on the client, so a session needs the credential's plaintext on that machine. See ADR 0001. Four roles, all of which are enforceable. Connect rides along with Read and is documented as a hint.
2FA ENFORCED and the per-member 2FA column server No two-factor concept exists anywhere — the only hit in the whole worktree is an aside in docs/crypto.md. Omitted. The member column carries what is known and matters: whether they have published a key a vault can be wrapped to.
LAST ACTIVE server UserAccount.LastSeenAtUtc is written at just-in-time provisioning and at enrollment and never on an ordinary authenticated request, so the column cannot answer "last active". Omitted.
Avatars server No picture is stored anywhere. Omitted; the row shows a name and an address.
Pending invites, resend, revoke server An invitation entity, a token with a lifetime, and an outbound mail path. MembershipStatus.Invited remains unwritten. Adding a member resolves an address the caller types against the directory, so the account has to have signed in here once. The screen says that when the lookup finds nothing.
SSO · OIDC · okta.dodotech.dev server Per-team SSO. Authentication is one global JWT scheme bound to one authority. Omitted.
A rekey after a membership change client Re-wrapping every item's data key under a fresh vault key, which only a client holding the current one can do. M5. The vault is flagged RekeyRequired and the row says a rotation is owed.
Ownership transfer server A confirmation flow and a rule for what happens to the outgoing owner. The owner cannot be removed or demoted, with its own problem code rather than a bare 400.

The trap this document warned about is still a trap. GET /api/v1/meta advertises features: ["teams"] unconditionally (MetaEndpoints.cs). It was meaningless when nothing implemented teams and it is meaningless now that everything does — it has never been computed from anything. Do not gate a client feature on it.


Preferences

The screen ships with what is real — this machine's device key, locking, and syncing — and lists the rest as absent rather than omitting it silently.

Design element Layer What it would take
Any preference, saved client-storage A seventh table in the client cache, or a preference item type in the vault. Which one matters: the vault syncs, so a preference stored there follows you between machines, and a terminal font size probably should not.
Terminal font, size, cursor style, cursor blink, scrollback client-terminal The store, and a frame to carry it. TerminalServerOpcode has four values (Output, SessionOpened, SessionClosed, SessionActivated) and none carries an option.
Font family picker offering IBM Plex Mono ui Shipping the font, as above.
Audible bell client-terminal + ui xterm.js 6 removed bellStyle and bellSound; what remains is an onBell event, so the page would have to make the sound itself.
GENERAL section ui There is no general setting to put in it. The theme is fixed by decision, and window size is not persisted.
KEYS & AGENT section client-ssh There is no agent, at all — no own agent, no forwarding, no Pageant or OpenSSH-agent interop.
SYNC & VAULT section client-domain Nothing here is adjustable. The auto-sync interval is a private static readonly with a remark arguing for its value.
SECURITY & SSO section server The SSO half needs team endpoints; there is no policy for the screen to show.
SHORTCUTS section ui There is no keybinding infrastructure and no rebinding surface.
Auto-lock after idle client-session An activity source, a decision about what counts as idle, and — the hard part — a policy for a shell mid-job. LockAsync's own remark already argues that an unattended timeout which killed a running job would be worse than the exposure it removed.
Require biometric to sign crypto Three separate falsehoods in one row. There is no signing service — the private key is decrypted and handed to SSH.NET whole, so there is no per-signature moment to interrupt; there are no connect-only keys; and per-use consent would need the key to live in the TPM, which is a different key hierarchy from the one in docs/crypto.md.
Pill toggle switches ui Purely presentational — App.axaml has no ToggleSwitch selector, so Fluent's default would appear un-themed beside the hairline chrome.

Two things the import changed on purpose

Hosts left the vault column. They have their own sidebar beside the terminal, and the vault screen holds keys, passwords and pinned host keys. This follows the design, and it is also the better split: the host list is what you look at while you work, and the rest is what you go and manage. VaultSection lost its Hosts member and gained All.

Tabs moved to the shell, not the vault. Locking disposes the vault and deliberately leaves shells running, so a tab list rebuilt per unlock would lose track of sessions that are still connected — the very sessions the unlock screen already counts. TerminalWorkspace gained SessionActivated on the wire, IsSessionLive, and a SessionEnded event so a tab can stop claiming to be connected.