mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-20 18:26:40 +01:00
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - Agent runs are interrupted when a server restarts, a deploy swaps
servers, a controller lease is lost, or a sandbox Runner cannot be
reached again
> - Exact recovery (resume the same session, replace, retry) refuses in
many ways, and each refusal leaves the task blocked or idle for a person
> - On staging, 577 of 1,000 runs in 11 hours failed or were
interrupted, across 25 error codes. About 900 earlier PRs each fixed one
refusal code, and the rate did not converge
> - This pull request adds one universal fallback: when an unfinished
run cannot be resumed exactly, for any reason that is not the person's,
the task continues in a fresh run from its last saved workspace and its
conversation
> - The benefit is that a new failure code next week falls back without
code changes, and tasks stop waiting on a person for infrastructure
faults
## Linked Issues or Issue Description
**What happened?**
After a server restart or deploy swap, many runs end with codes such as
`native_continuation_requires_reconciliation`
(`runner_remote_process_identity_unavailable`), `controller_lease_lost`,
`process_lost`, `server_shutdown_interrupted` and
`native_session_interrupted` (including `runner_harness_state_mismatch:
resumed_not_suspended` / `resumed_unavailable`). The disposition worker
then settles the recovery hold as "blocked, no replay". The task stays
Blocked until a person presses Retry or reconciles the run.
**Expected behavior**
A task waits on a person only for reasons that are the person's: an
approval, a question, a credential to share, a budget stop, a provider
usage limit, or a Stop they asked for. For every other failure, the
system starts a new run for the same task from the last verified saved
workspace, without a person.
**Steps to reproduce**
1. Start a task on a sandbox agent.
2. Restart or swap the server while the run is in progress, so the new
server cannot re-reach the Runner.
3. See that the run fails with
`native_continuation_requires_reconciliation` and the task stays Blocked
with no successor run.
**Paperclip version or commit**
`master` at 9e910ff9ae.
## What Changed
- New module `server/src/services/fresh-continuation-fallback.ts`. One
entry point, `ensureFreshContinuation` / `continueFreshInTransaction`,
takes a run that ended without finishing and guarantees one of:
something already follows it (a newer run, a linked replacement, an
operator decision), exact recovery is still under way (checked again
later), the cause is the person's, or one fresh continuation is
recorded.
- Classification, not an allowlist: `PERSON_OWNED_FAILURE_CODES` plus
task-level waits (pending approval, pending question, paused or
unavailable agent, paused company, review in progress, a Stop that a
person asked for, a Blocked status that another decision owns). Every
other code, including unknown future codes, falls back.
- Idempotent per source run. The task row lock serializes callers on
every server; the recorded `freshContinuation.sourceRunId` makes later
calls a no-op.
- Isolation, never waiting on proof. A local process still running on
this host is settled by its own owner first. If the old local writer is
proven gone, the task workspace is kept. Otherwise the continuation gets
a new isolated workspace (`fresh_isolated_workspace`), so two runs never
write one directory. The decision records the stop proof honestly
(`providerStopProof: isolated`).
- Claim admission no longer waits on the cleanup of a predecessor that a
fresh continuation replaced (`heartbeat/queue.ts`, predecessor loop).
Without this, the successor stayed queued behind the unreachable lease,
which is the staging failure itself.
- Only reconciliation holds that this decision owns are settled.
Workspace restore holds that workspace recovery still owns, and resolved
history, are untouched.
- `continuation_user_authorization_missing` is the system's failure (a
Retry's authorization was not carried forward), not the person's.
- Safety: the old run keeps its fencing. When it still owns its remote
workspace epoch, the epoch is abandoned
(`abandonUnverifiedRemoteWorkspaceEpoch`), so the old writer can never
publish and the next run restores the last committed save into a new
root on a new allocation. The unverified files stay in the abandoned
root, unapplied. The native coordinator records `replacementDenied:
fresh_continuation`, so no exact replay also starts.
- The replaced run loses its write authority. The fallback records
`freshContinuationSupersededAt` on the old run. The agent-JWT middleware
(`loadAgentRunAuthentication` reads one more field in its existing row
read) and `assertAgentRunWriteAllowed` refuse every write of the old
token with `agent_run_superseded`. Before this change, a failed or
interrupted run's token could still write. So an agent that is still
alive in an unreachable sandbox cannot act on the task next to its
continuation.
- The opening context (`recoveryOutcomes` in the execution continuation)
says that the previous attempt was cut off and lists the actions with
unverified outcomes, so the agent checks before it repeats a side
effect. No copy asks a person to inspect or repair a sandbox, workspace
or run. The hold text written at run end no longer says "Inspect the
original failure and reconcile".
- Delivery reuses the existing path: the settled hold gets an
`executionReconciliation` decision and `continuationDelivery: pending`,
and `deliverReconciledExecutions` starts the run (fresh session,
`retryOfRunId` = source).
- Budget per task, not per restart: at most 6 fresh continuations per
task per rolling hour (the server-restart lane's cap). At the cap the
hold stays with the system and is checked again when the window has
room. The failure allowance is never charged.
- Call sites:
- `settleUnrecoverableExecutions` (execution-recovery-resolution.ts):
every hold whose exact recovery is ruled out now goes through the
fallback before it would block the task. Interrupted preparations keep
their own replay lane first.
- `continueHeldExecutions` (same worker): picks up holds saved by older
releases (settled-but-blocked no-replay holds, finished or exhausted
saved-work recoveries) for the task's latest unfinished run.
- Wake admission (heartbeat/queue.ts): an ordinary system continuation
wake (sweep, monitor, timer) that a hold would defer now calls the
fallback in the same transaction and becomes the fresh continuation
itself. This is the path the idle-task sweep uses. A person's Retry,
message or queued send keeps its exact admission; the disposition worker
settles the hold, and the saved input then continues.
- `executionControlSchedule.disposition` wakes the worker for held tasks
and for the next due recheck.
### Human gate versus system uncertainty
One documented rule (`isHumanGate` plus `personOwnedTaskWait`) decides
what the fallback never touches. The rule goes by cause, never by owner
type:
- Human gates: a pending approval, a pending question, a user-owned
hold, an owner-gone attestation, a recorded explicit continuation, an
operator reconciliation, a person's Stop.
- System uncertainty: a no-replay record ("outcome unknown after an
interruption"), a hold that is `board`-owned only by default, an
unverifiable controller process space or host.
For uncertainty the fallback starts a fresh run, never a replay. The
interrupted turn is not re-sent. No tool call or write is re-executed.
Governed actions go through their approval gate again. The earlier
`automaticRecovery` decision is kept on the hold as
`automaticRecoveryHistory`, beside the attributed
`freshContinuationSettlement` and an activity entry.
### Restart recovery refused for an unverifiable controller
The case is `controller_process_space_unverified`, its remote variant,
or `controller_host_unrecorded` / `controller_host_unprovable`. Its
lease has lapsed; a live, renewing lease is still waited for (server
overlap). After `NATIVE_NAMESPACE_UNVERIFIED_WAIT_MS` (2 minutes) the
run ends (failed, `native_restart_recovery_blocked`). That frees the
task's execution lock and the agent's slot, and a system-owned hold lets
the fallback continue the task. Recording the stops is optional. The "a
board operator must record both stops" text is replaced.
Board/operator attention sites in native restart recovery:
- `surfaceNativeControllerNamespaceAttention`: copy replaced; the run
now ends after the bounded wait.
- `requestNativeControllerStopConfirmation`
(native-controller-host-recovery.ts): copy says confirmation is
optional; the same bounded end applies.
- The blocked restart-recovery hold (`nextAction` "Inspect the preserved
checkpoint and reconcile…"): copy replaced; the fallback settles the
hold by cause.
- `owner-gone-attestation.ts` (generic resource records from older
releases): unchanged. It is not a run-recovery path, and it is listed as
a follow-up.
### Every wait and its bound
| Wait | Owner that rechecks | Worst-case total before the fresh
continuation |
| --- | --- | --- |
| `exact_recovery_pending` (execution settling, native finalizer,
controller lease, interrupted-preparation lane, saved-work recovery,
restore hold, retained-source fallback, live local process) |
Disposition worker, woken at each deadline; rechecks at 1, 2, 3 and 4
minutes, then every 4 | About 10 minutes from the first pending check.
Up to 15 while the owner shows progress (a new event, output, lease
renewal or coordinator update since the last check). |
| Saved-work (workspace) recovery that is `scheduled`,
`awaiting_evidence`, or `running` without progress | Held-run scan, now
selected and scheduled like the other holds | About 10 minutes; 15 at
most |
| Saved-work recovery that is `running` and advancing (revision,
attempt, state or hold update since the last check) | Held-run scan |
Its own 30-minute budget |
| Unverifiable lapsed controller in restart recovery | Restart recovery
(`retryAt`) | 2 minutes, then the run ends and the fallback continues it
within one disposition pass |
| Hourly cap (`backoff`) | Disposition worker, woken at the deadline |
Until the rolling hour frees a slot, at most 1 hour. This is the
deliberate restart-lane bound, not a stuck state. |
| Person-owned or dependency wait | The person's own action wakes the
task; rechecked hourly | Not bounded: it is the person's |
Restart-recovery classifications (`claimNativeRestartRecoveries`, which
`recoverNativeRunsAfterRestart` calls):
| Classification | Who rechecks, when | Bound |
| --- | --- | --- |
| Claims: `reattach_existing_runner`, `reattach_remote_runner`,
`resume_dead_runner`, `bootstrap_incomplete` | The claiming server
resumes exactly | Exact resume; no fallback |
| `already_finalized` | Nobody; it is terminal | None needed |
| `blocked` | The run is terminal and its hold goes to the fallback
(disposition worker) | Fallback bounds above |
| `awaiting_evidence` / `current_controller_preparing` | This process's
own executor | Ends with that executor |
| `awaiting_evidence` / `recovery_rows_busy` | Restart dispatch after 1
second | Self-resolving: the row lock clears |
| `awaiting_evidence` / a live lease (`controller_lease_held`,
`controller_lease_grace`, `controller_still_alive`) | Restart dispatch
at the lease-renew interval | Never ended while the old controller
renews its lease |
| `awaiting_evidence` / unverifiable lapsed controller (process space or
host) | Restart dispatch every 2 minutes | Ends 2 minutes after the
lease lapsed |
| `awaiting_evidence` / any other reason (startup cleanup, finalizer or
cleanup owner, recovery failure, routine busy, future reasons) | Restart
dispatch at the owner's `retryAt`, or within 2 minutes when it has none
| Ends 2 minutes after the lease lapsed, or after 10 minutes, whichever
comes first. It never ends while a lease is live. |
Ending a run sets it failed (`native_restart_recovery_blocked`),
releases the task lock and the agent's slot, and records a system-owned
hold. The fallback then continues the task once, isolated.
No wait needs longer. Saved-work recovery and retained-source checks
keep their own attempt budgets. After 10 to 15 minutes the fallback
continues anyway in a new root, without the unverified files, which stay
recorded and unapplied.
### When the fallback preempts saved-work recovery
The unverified files are not destroyed. They stay where recovery found
them: in the abandoned remote epoch's root, or in the old local
workspace that the task is no longer bound to. They are recorded and
never applied. Settling the hold records the decision on it, so the
retained-workspace worker stops scheduling it (its query requires no
`executionReconciliation` unless recovery already succeeded). If an
attempt already in flight finishes later, it cannot publish into the
fresh run's workspace. For a remote workspace, the epoch it targets is
abandoned and the writer is cleared. For a local one, the task now
points at a new workspace. Whenever a wait is preempted, the
continuation is isolated: a fenced remote epoch or a new local
workspace.
### What people see
The task's interruption presentation shows "working" while the primary
fresh-continuation record is pending and has not started (coalesced
holds never count, and they follow the primary's final state), and a
timed wait with no Retry while a check is scheduled. The Retry banner
appears only when the system has given up or the cause is the person's.
Holds settled together with the primary record are `coalesced`; they no
longer show "closed or reassigned". A pending continuation is never a
second run: if a person's Retry or saved input started a run after it
was recorded, delivery links to that run instead. A task that finished
before delivery gets no run.
### Readers of what the fallback writes
- Recheck deadline (`freshContinuationCheck`):
- `unresolvedExecutionDispositions` and
`heldExecutionsAwaitingFreshContinuation` read it to skip work that is
not due.
- `nextFreshContinuationCheckAt` and
`executionControlSchedule.disposition` read it to schedule the worker.
- Tests: the cap deadline is scheduled, and a cancelled task goes quiet.
- Isolation decision:
- The task's workspace binding is cleared. Every path that starts the
next run (system continuation, saved input, Retry, sweep wake) resolves
the workspace from that binding at run start (heartbeat.ts
`issueContext.executionWorkspaceId`).
- `freshContinuationIsolation` in the run context is read by
`readFreshWorkspaceReadiness` for the shared-directory case.
- `freshContinuationSupersededAt` on the old run is read by the auth
middleware, `assertAgentRunWriteAllowed` and the claim gate.
## Verification
- `pnpm vitest run src/services/fresh-continuation-fallback.test.ts`
(server): every test passes (it covers the cases below).
- The continuation starts (leaves the queue) while the unreachable old
lease is `active` or `pending_cleanup`. This was red before the
claim-admission fix.
- Isolation: a writer on another host gets a new isolated workspace; a
process alive on this host waits for its owner; a proven-gone writer
keeps the workspace.
- Owned holds only: a restore hold and resolved history are untouched.
- The recheck schedule goes quiet when the task is cancelled.
- `continuation_user_authorization_missing` continues under the task's
responsible user.
- One test per top-five staging code (native and legacy hold shapes):
the task gets exactly one fresh successor run from saved state.
- `runner_harness_state_mismatch` (`resumed_not_suspended`,
`resumed_unavailable`) falls back.
- An unknown code falls back.
- Two concurrent callers plus the disposition worker create one
successor.
- Old-saved/new-resumed: a settled-but-blocked hold from the previous
release, with its coordinator in `awaiting_runner_reattach`, continues
on the first pass.
- An ordinary system continuation wake is admitted as the fresh
continuation.
- Fencing: after the continuation is recorded, the old run's token is
refused for writes (`agent_run_superseded`) and the continuation's token
is not. The old run's remote workspace epoch is abandoned: the writer is
cleared, the location and provider root are cleared, and the epoch
advances. So the next preparation restores the last save into a new root
(`prepareRemoteWorkspace`: `sameAllocation` is false when the location
is null, server/src/services/remote-workspaces.ts:437-470).
- The opening context names the cut-off run and an unverified tool
action.
- Person-owned causes (pending approval, credential not shared, budget
stop) are not continued.
- The hourly cap backs off without blocking, then continues.
- Red/green: with the wiring reverted to `master`, 9 of these tests fail
(the task stays Blocked with no successor).
- Affected suites pass locally: `execution-recovery-dead-controller`,
`completed-turn-cleanup-hold-settle`, `execution-control-work`,
`heartbeat-process-recovery`, `issue-recovery-actions`,
`native-safe-replacement`, `native-failed-workspace-recovery`,
`task-retry`, `runner-lost-mid-turn-continuation`,
`workspace-recovery-fallback`, `explicit-native-continuation`,
wake-queue `postgres`, `agent-conversations`, `native-startup-retry`,
`governed-wait-runner-loss`, `native-preclaim-continuation`.
- `pnpm verify --explain` selects the full suite (the wake queue is
shared), so CI supplies full coverage.
Existing expectations changed because they asserted the "wait for a
person" behavior this PR replaces:
- `issue-monitor-scheduler.test.ts`: "explains held monitor dispatch
without replaying uncertain execution" is now "continues a held task in
a fresh run when its monitor is due". A due monitor on a task held by a
`process_lost` hold now starts one fresh run instead of deferring.
- `agent-run-authentication.test.ts`: the auth row also reads
`freshContinuationSupersededAt` (still under 512 bytes), and a
superseded run is refused.
- `assigned-work-invariant.test.ts` (#21139): "with the real wake queue,
a hold that defers the continuation leaves a violation…" is now "…a hold
reached by the sweep's wake becomes one pending fresh continuation". The
sweep's wake now yields one pending fresh continuation, and the next
sweep adds no attempt.
- `heartbeat-native-restart-suspension.test.ts` and
`native-runner-restart-recovery.integration.test.ts`: awaiting-evidence
dispositions now carry a `retryAt`, because every wait is rechecked. A
lapsed unverifiable controller no longer waits forever.
- `native-remote-retry-admission.test.ts`: the blocked-hold text is now
"This run stopped before it finished. Your saved work is kept." The old
text asked people to reconcile.
Holdless legacy runs: in the 1,000 staging runs, 0 tasks end on
`controller_lease_lost` or `process_lost` (every one has a later run).
12 end on `server_shutdown_interrupted` with no retry scheduled (attempt
0). On staging now, 9 of those 12 sit `in_progress` with no run and no
hold, 1 is done and 2 have a live run. Those 9 belong to the idle-task
sweep (#21139).
## Risks
- Speed: nothing is added to the normal path before `turn.start`. The
fallback runs only on the disposition worker after exact recovery is
ruled out, or at wake admission when a hold already defers the wake. The
auth middleware reads one more JSON field in the row read it already
does.
- Behavior shift: tasks that today stay Blocked for a person after an
infrastructure failure now continue automatically. A side effect with an
unknown outcome is not replayed; the new run is told about it and must
check first.
- After deploy, the disposition worker continues tasks already blocked
by these codes, at most 25 per pass and at most 6 per task per hour. No
migration. No manual Retry is needed.
- The old executor is not proven stopped before the successor starts.
Fencing is: its API writes are refused, its workspace epoch is
abandoned, the native `replacementDenied` marker is set, and the
successor has a separate run identity.
Known gaps (follow-ups):
- No quarantine attachment: unverified files stay unapplied in the
abandoned workspace root (recorded on the workspace in
`remoteWorkspaceAbandonedEpochs`). They are not read into a task
attachment, because the sandbox is usually unreachable in these
failures.
- Agent Chat conversations are excluded at wake admission. They keep
their owner-bound continuation.
- Under a `reuse_by_environment` lease policy the new allocation can be
on the same provider sandbox. It always uses a new root and a new epoch,
and the old run cannot publish or write.
- Tasks left `in_progress` with no run and no hold (for example after
`server_shutdown_interrupted` with no retry scheduled) belong to the
idle-task sweep. That sweep's wake reaches this fallback when a hold
exists.
> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.
## Model Used
- Claude Opus 5.5 (`claude-opus-5-5`), Anthropic, via Claude Code with
tool use (shell, file edits, test runs).
## Checklist
- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [ ] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
1042 lines
39 KiB
TypeScript
1042 lines
39 KiB
TypeScript
import { notifyMuseContactVerification } from "../modules/external-agents/index.js";
|
|
import { idleApiActivity } from "../services/idle-api-activity.js";
|
|
import { createHash, timingSafeEqual } from "node:crypto";
|
|
import type { Request, RequestHandler } from "express";
|
|
import { and, eq, isNull, sql } from "drizzle-orm";
|
|
import type { Db } from "@paperclipai/db";
|
|
import {
|
|
activityLog,
|
|
agentApiKeys,
|
|
agents,
|
|
authUsers,
|
|
companies,
|
|
companyMemberships,
|
|
heartbeatRuns,
|
|
instanceUserRoles,
|
|
} from "@paperclipai/db";
|
|
import {
|
|
MAX_ISSUE_PREFIX_ATTEMPTS,
|
|
deriveIssuePrefixBase,
|
|
isIssuePrefixConflict,
|
|
issuePrefixSuffixForAttempt,
|
|
pickAvailableIssuePrefix,
|
|
rekeyCompanyIssueIdentifiers,
|
|
} from "../services/issue-prefix.js";
|
|
import { verifyLocalAgentJwt } from "../agent-auth-jwt.js";
|
|
import { agentRunSupersededByFreshContinuation, agentRunWritesRevoked } from "../agent-run-cancellation.js";
|
|
import { loadAgentRunAuthentication } from "./agent-run-authentication.js";
|
|
import { isUuidLike, normalizeAgentApiKeyScope, type DeploymentMode } from "@paperclipai/shared";
|
|
import type { BetterAuthSessionResult } from "../auth/better-auth.js";
|
|
import { logger } from "./logger.js";
|
|
import { captureRunIdentity } from "../services/run-identity.js";
|
|
import { boardAuthService } from "../services/board-auth.js";
|
|
import { retryIdempotentDatabaseOperation } from "../database-retry.js";
|
|
import { beginIdleTrackedWork } from "../services/task-admission.js";
|
|
import { notifyChatCompanyAuthorityChange } from "../services/chat-run-milestone-notifications.js";
|
|
|
|
export {
|
|
isTransientDbConnectionError,
|
|
retryIdempotentDatabaseOperation as retryOnTransientDbConnectionError,
|
|
} from "../database-retry.js";
|
|
|
|
import { syncCloudTenantWrites } from "./cloud-tenant-sync.js";
|
|
import { resolveOwnerInstanceAdmin } from "../services/cloud-tenant-authority.js";
|
|
import { ensureHumanRoleDefaultGrants } from "../services/principal-access-compatibility.js";
|
|
import { forbidden, unauthorized, unprocessable } from "../errors.js";
|
|
|
|
export { isCloudManagedInstance } from "../services/cloud-instance.js";
|
|
import { cloudTenantPrimaryCompanyId, isCloudManagedInstance } from "../services/cloud-instance.js";
|
|
|
|
function hashToken(token: string) {
|
|
return createHash("sha256").update(token).digest("hex");
|
|
}
|
|
|
|
function normalizeOptionalString(value: string | null | undefined) {
|
|
return value?.trim() || null;
|
|
}
|
|
|
|
function invalidAgentTokenMessage(token: string) {
|
|
try {
|
|
const payload = JSON.parse(Buffer.from(token.split(".")[1] ?? "", "base64url").toString("utf8")) as {
|
|
exp?: unknown;
|
|
};
|
|
if (typeof payload.exp === "number" && payload.exp <= Math.floor(Date.now() / 1000)) {
|
|
return "Expired agent token; obtain fresh credentials and retry";
|
|
}
|
|
} catch {
|
|
// Malformed and incorrectly signed tokens share the generic failure below.
|
|
}
|
|
return "Agent token did not verify; obtain fresh credentials and retry";
|
|
}
|
|
|
|
async function resolveLegacyRunResponsibleUserId(
|
|
db: Db,
|
|
input: { companyId: string; agentId: string; runId: string },
|
|
) {
|
|
if (!isUuidLike(input.runId)) return null;
|
|
const run = await db
|
|
.select({ responsibleUserId: heartbeatRuns.responsibleUserId })
|
|
.from(heartbeatRuns)
|
|
.where(
|
|
and(
|
|
eq(heartbeatRuns.id, input.runId),
|
|
eq(heartbeatRuns.companyId, input.companyId),
|
|
eq(heartbeatRuns.agentId, input.agentId),
|
|
),
|
|
)
|
|
.then((rows) => rows[0] ?? null);
|
|
return normalizeOptionalString(run?.responsibleUserId);
|
|
}
|
|
|
|
async function loadResponsibleUserMemberships(
|
|
db: Db,
|
|
input: { companyId: string; userId: string | null },
|
|
) {
|
|
if (!input.userId) return [];
|
|
const [user, memberships] = await Promise.all([
|
|
db
|
|
.select({ id: authUsers.id })
|
|
.from(authUsers)
|
|
.where(eq(authUsers.id, input.userId))
|
|
.then((rows) => rows[0] ?? null),
|
|
db
|
|
.select({
|
|
companyId: companyMemberships.companyId,
|
|
membershipRole: companyMemberships.membershipRole,
|
|
status: companyMemberships.status,
|
|
})
|
|
.from(companyMemberships)
|
|
.where(
|
|
and(
|
|
eq(companyMemberships.companyId, input.companyId),
|
|
eq(companyMemberships.principalType, "user"),
|
|
eq(companyMemberships.principalId, input.userId),
|
|
eq(companyMemberships.status, "active"),
|
|
),
|
|
),
|
|
]);
|
|
return user ? memberships : [];
|
|
}
|
|
|
|
/**
|
|
* The user's own active company memberships — the exact company scope a
|
|
* locally authenticated session actor carries. Shared by the session path
|
|
* and the Cloud trusted-header path so both resolve the same access set.
|
|
*/
|
|
async function loadActiveUserCompanyMemberships(db: Db, userId: string) {
|
|
return db
|
|
.select({
|
|
companyId: companyMemberships.companyId,
|
|
membershipRole: companyMemberships.membershipRole,
|
|
status: companyMemberships.status,
|
|
})
|
|
.from(companyMemberships)
|
|
.where(
|
|
and(
|
|
eq(companyMemberships.principalType, "user"),
|
|
eq(companyMemberships.principalId, userId),
|
|
eq(companyMemberships.status, "active"),
|
|
),
|
|
);
|
|
}
|
|
|
|
async function auditAgentJwtRunHeaderMismatch(
|
|
db: Db,
|
|
input: { companyId: string; agentId: string; claimRunId: string; headerRunId: string; method: string; url: string },
|
|
) {
|
|
try {
|
|
// The claimed run id is untrusted input from a rejected token. activity_log
|
|
// .run_id is a real FK to heartbeat_runs, so writing it unconditionally
|
|
// makes the insert fail — and the catch swallows it — exactly when the
|
|
// claim is bogus. Keep the claim in entityId/details and populate the FK
|
|
// column only when the row actually exists.
|
|
const claimedRunExists = isUuidLike(input.claimRunId)
|
|
? await db
|
|
.select({ id: heartbeatRuns.id })
|
|
.from(heartbeatRuns)
|
|
.where(eq(heartbeatRuns.id, input.claimRunId))
|
|
.then((rows) => rows.length > 0)
|
|
: false;
|
|
await db.insert(activityLog).values({
|
|
companyId: input.companyId,
|
|
actorType: "agent",
|
|
actorId: input.agentId,
|
|
action: "auth.agent_jwt_run_header_mismatch",
|
|
entityType: "heartbeat_run",
|
|
entityId: input.claimRunId,
|
|
...(isUuidLike(input.agentId) ? { agentId: input.agentId } : {}),
|
|
...(claimedRunExists ? { runId: input.claimRunId } : {}),
|
|
details: {
|
|
claimRunId: input.claimRunId,
|
|
headerRunId: input.headerRunId,
|
|
method: input.method,
|
|
url: input.url,
|
|
},
|
|
});
|
|
} catch (err) {
|
|
logger.warn(
|
|
{ err, companyId: input.companyId, agentId: input.agentId, claimRunId: input.claimRunId },
|
|
"Failed to audit rejected agent JWT run header mismatch",
|
|
);
|
|
}
|
|
}
|
|
|
|
async function auditAgentKeyMissingResponsibleUser(
|
|
db: Db,
|
|
input: { companyId: string; agentId: string; keyId: string; method: string; url: string },
|
|
) {
|
|
try {
|
|
await db.insert(activityLog).values({
|
|
companyId: input.companyId,
|
|
actorType: "agent",
|
|
actorId: input.agentId,
|
|
action: "auth.agent_key_missing_responsible_user",
|
|
entityType: "agent_api_key",
|
|
entityId: input.keyId,
|
|
...(isUuidLike(input.agentId) ? { agentId: input.agentId } : {}),
|
|
details: {
|
|
method: input.method,
|
|
url: input.url,
|
|
},
|
|
});
|
|
} catch (err) {
|
|
logger.warn(
|
|
{ err, companyId: input.companyId, agentId: input.agentId, keyId: input.keyId },
|
|
"Failed to audit rejected agent key without responsible user binding",
|
|
);
|
|
}
|
|
}
|
|
|
|
interface ActorMiddlewareOptions {
|
|
deploymentMode: DeploymentMode;
|
|
resolveSession?: (req: Request) => Promise<BetterAuthSessionResult | null>;
|
|
}
|
|
|
|
const publicRoutineWebhookPath = /^\/api\/routine-triggers\/public\/[a-f0-9]{24}\/fire\/?$/i;
|
|
|
|
const publicMcpGatewayProtocolPath = /^\/mcp\/gateways\/gw_[a-f0-9]{32}\/?$/i;
|
|
|
|
export function actorMiddleware(db: Db, opts: ActorMiddlewareOptions): RequestHandler {
|
|
const boardAuth = boardAuthService(db);
|
|
const authenticate: RequestHandler = async (req, _res, next) => {
|
|
req.actor =
|
|
opts.deploymentMode === "local_trusted"
|
|
? {
|
|
type: "board",
|
|
userId: "local-board",
|
|
userName: "Local Board",
|
|
userEmail: null,
|
|
isInstanceAdmin: true,
|
|
source: "local_implicit",
|
|
}
|
|
: { type: "none", source: "none" };
|
|
|
|
// Routine ingress authenticates its own bearer/signature. Never interpret
|
|
// webhook credentials as agent keys or attach an ambient browser session.
|
|
if (req.method === "POST" && publicRoutineWebhookPath.test(req.path)) {
|
|
req.actor = { type: "none", source: "none" };
|
|
next();
|
|
return;
|
|
}
|
|
|
|
const runIdHeader = req.header("x-paperclip-run-id");
|
|
|
|
const authHeader = req.header("authorization");
|
|
const hasBearerCredentials = /^bearer(?:\s|$)/i.test(authHeader ?? "");
|
|
|
|
// Public MCP gateway protocol requests carry a pcgw_* bearer that is
|
|
// validated by the gateway service itself. Do not interpret that bearer as
|
|
// a board key or agent JWT here: doing so rejects the MCP handshake before
|
|
// the protocol route can verify its run-scoped credential. Keep this bypass
|
|
// restricted to the unguessable public gateway path; all /api routes retain
|
|
// the normal actor authentication path below.
|
|
if (hasBearerCredentials && publicMcpGatewayProtocolPath.test(req.path)) {
|
|
if (runIdHeader) req.actor.runId = runIdHeader;
|
|
next();
|
|
return;
|
|
}
|
|
|
|
if (!hasBearerCredentials) {
|
|
if (opts.deploymentMode === "authenticated" && opts.resolveSession) {
|
|
const cloudTenantActor = await resolveCloudTenantActor(db, req);
|
|
if (cloudTenantActor) {
|
|
req.actor = {
|
|
...cloudTenantActor,
|
|
runId: runIdHeader ?? undefined,
|
|
};
|
|
next();
|
|
return;
|
|
}
|
|
|
|
let session: BetterAuthSessionResult | null = null;
|
|
try {
|
|
session = await opts.resolveSession(req);
|
|
} catch (err) {
|
|
logger.warn(
|
|
{ err, method: req.method, url: req.originalUrl },
|
|
"Failed to resolve auth session from request headers",
|
|
);
|
|
}
|
|
if (session?.user?.id && session.session?.id) {
|
|
const userId = session.user.id;
|
|
const [roleRow, memberships] = await Promise.all([
|
|
db
|
|
.select({ id: instanceUserRoles.id })
|
|
.from(instanceUserRoles)
|
|
.where(and(eq(instanceUserRoles.userId, userId), eq(instanceUserRoles.role, "instance_admin")))
|
|
.then((rows) => rows[0] ?? null),
|
|
loadActiveUserCompanyMemberships(db, userId),
|
|
]);
|
|
req.actor = {
|
|
type: "board",
|
|
userId,
|
|
sessionId: session.session.id,
|
|
userName: session.user.name ?? null,
|
|
userEmail: session.user.email ?? null,
|
|
companyIds: memberships.map((row) => row.companyId),
|
|
memberships,
|
|
isInstanceAdmin: Boolean(roleRow),
|
|
runId: runIdHeader ?? undefined,
|
|
source: "session",
|
|
};
|
|
next();
|
|
return;
|
|
}
|
|
}
|
|
if (runIdHeader) req.actor.runId = runIdHeader;
|
|
next();
|
|
return;
|
|
}
|
|
|
|
const token = authHeader!.slice("bearer".length).trim();
|
|
if (!token) {
|
|
next(unauthorized("Empty bearer token; provide valid agent credentials and retry"));
|
|
return;
|
|
}
|
|
|
|
const boardKey = await boardAuth.findBoardApiKeyByToken(token);
|
|
if (boardKey) {
|
|
const access = await boardAuth.resolveBoardAccess(boardKey.userId);
|
|
if (access.user) {
|
|
await boardAuth.touchBoardApiKey(boardKey.id);
|
|
req.actor = {
|
|
type: "board",
|
|
userId: boardKey.userId,
|
|
userName: access.user?.name ?? null,
|
|
userEmail: access.user?.email ?? null,
|
|
companyIds: access.companyIds,
|
|
memberships: access.memberships,
|
|
isInstanceAdmin: access.isInstanceAdmin,
|
|
keyId: boardKey.id,
|
|
runId: runIdHeader || undefined,
|
|
source: "board_key",
|
|
};
|
|
if (isCloudManagedInstance() && (access.isInstanceAdmin || access.companyIds.length)) {
|
|
_res.once("finish", () => { if (_res.statusCode >= 200 && _res.statusCode < 400) idleApiActivity.record(); });
|
|
}
|
|
next();
|
|
return;
|
|
}
|
|
}
|
|
|
|
const tokenHash = hashToken(token);
|
|
const key = await db
|
|
.select()
|
|
.from(agentApiKeys)
|
|
.where(and(eq(agentApiKeys.keyHash, tokenHash), isNull(agentApiKeys.revokedAt)))
|
|
.then((rows) => rows[0] ?? null);
|
|
|
|
if (!key) {
|
|
const claims = verifyLocalAgentJwt(token);
|
|
if (!claims) {
|
|
next(unauthorized(invalidAgentTokenMessage(token)));
|
|
return;
|
|
}
|
|
|
|
const agentRecord = await db
|
|
.select()
|
|
.from(agents)
|
|
.where(eq(agents.id, claims.sub))
|
|
.then((rows) => rows[0] ?? null);
|
|
|
|
if (!agentRecord || agentRecord.companyId !== claims.company_id) {
|
|
next(unauthorized("Agent record is missing or belongs to another company; obtain fresh credentials and retry"));
|
|
return;
|
|
}
|
|
|
|
if (agentRecord.lifecycleHolds?.includes("setup")) {
|
|
next(unauthorized("Agent setup is incomplete and cannot authenticate"));
|
|
return;
|
|
}
|
|
if (agentRecord.status === "terminated") {
|
|
next(unauthorized("Agent is terminated and cannot authenticate"));
|
|
return;
|
|
}
|
|
if (agentRecord.status === "pending_approval") {
|
|
next(unauthorized("Agent is pending approval and cannot authenticate"));
|
|
return;
|
|
}
|
|
|
|
const normalizedRunIdHeader = normalizeOptionalString(runIdHeader);
|
|
if (normalizedRunIdHeader && normalizedRunIdHeader !== claims.run_id) {
|
|
await auditAgentJwtRunHeaderMismatch(db, {
|
|
companyId: claims.company_id,
|
|
agentId: claims.sub,
|
|
claimRunId: claims.run_id,
|
|
headerRunId: normalizedRunIdHeader,
|
|
method: req.method,
|
|
url: req.originalUrl,
|
|
});
|
|
next(
|
|
unprocessable("X-Paperclip-Run-Id does not match signed agent JWT run_id", {
|
|
code: "agent_jwt_run_id_mismatch",
|
|
claimRunId: claims.run_id,
|
|
headerRunId: normalizedRunIdHeader,
|
|
}),
|
|
);
|
|
return;
|
|
}
|
|
|
|
const [identityRun] = await loadAgentRunAuthentication(db, {
|
|
runId: claims.run_id,
|
|
companyId: claims.company_id,
|
|
agentId: claims.sub,
|
|
});
|
|
if (agentRunWritesRevoked(identityRun)
|
|
&& !["GET", "HEAD", "OPTIONS"].includes(req.method)) {
|
|
const conversation = identityRun?.contextSnapshot?.conversationMode === true;
|
|
if (agentRunSupersededByFreshContinuation(identityRun)) {
|
|
_res.status(403).json({ error: "A newer run continues this task. End the turn.", code: "agent_run_superseded" });
|
|
return;
|
|
}
|
|
_res.status(403).json({ error: conversation ? "This conversation turn was cancelled" : "This run was cancelled",
|
|
code: conversation ? "conversation_turn_cancelled" : "agent_run_cancelled" });
|
|
return;
|
|
}
|
|
if (identityRun?.activeIdentityContextId && identityRun.status === "running") {
|
|
const captured = await captureRunIdentity(db, { companyId: claims.company_id, agentId: claims.sub, runId: claims.run_id },
|
|
{ readScope: "request_authentication" });
|
|
identityRun.activeIdentityContextId = captured.context?.id ?? null;
|
|
identityRun.responsibleUserId = captured.context?.responsibleUserId ?? null;
|
|
}
|
|
if (claims.authorizing_user_id) {
|
|
// This signed projection exists only for internal native tools. Recheck
|
|
// the admitted run and current immutable binding on every API call.
|
|
try {
|
|
const { parseNativeExecutionInput } = await import("../vendor/paperclip-runner/index.js");
|
|
const { museRunnerBroker } = await import("../services/muse-runner-broker.js");
|
|
const [nativeRun] = await db.select({ status: heartbeatRuns.status, runtimeMode: heartbeatRuns.runtimeMode,
|
|
nativeExecutionInput: sql<unknown>`${heartbeatRuns.runnerProfileJson}->'nativeExecutionInput'`,
|
|
}).from(heartbeatRuns).where(and(
|
|
eq(heartbeatRuns.id, claims.run_id), eq(heartbeatRuns.companyId, claims.company_id), eq(heartbeatRuns.agentId, claims.sub),
|
|
));
|
|
const execution = parseNativeExecutionInput(nativeRun?.nativeExecutionInput);
|
|
if (nativeRun?.status !== "running" || nativeRun.runtimeMode !== "native" || execution.schema !== "paperclip.native-execution-input.v8"
|
|
|| execution.binding.runId !== claims.run_id || execution.binding.agentId !== claims.sub || execution.binding.companyId !== claims.company_id
|
|
|| await museRunnerBroker(db).authorizingUserIdForBinding(execution.provider.binding) !== claims.authorizing_user_id) {
|
|
throw new Error("Muse internal tool authority mismatch");
|
|
}
|
|
await museRunnerBroker(db).assertRunAuthority(execution);
|
|
} catch {
|
|
_res.status(403).json({ error: "Muse native tool authority is unavailable" });
|
|
return;
|
|
}
|
|
}
|
|
const onBehalfOfUserId = identityRun?.activeIdentityContextId
|
|
? identityRun.responsibleUserId
|
|
: claims.responsible_user_id !== undefined
|
|
? normalizeOptionalString(claims.responsible_user_id)
|
|
: await resolveLegacyRunResponsibleUserId(db, {
|
|
companyId: claims.company_id,
|
|
agentId: claims.sub,
|
|
runId: claims.run_id,
|
|
});
|
|
const onBehalfOfMemberships = await loadResponsibleUserMemberships(db, {
|
|
companyId: claims.company_id,
|
|
userId: onBehalfOfUserId,
|
|
});
|
|
|
|
req.actor = {
|
|
type: "agent",
|
|
agentId: claims.sub,
|
|
companyId: claims.company_id,
|
|
keyId: undefined,
|
|
keyScope: normalizeAgentApiKeyScope(claims.key_scope),
|
|
runId: claims.run_id,
|
|
onBehalfOfUserId,
|
|
...(claims.authorizing_user_id ? { authorizingUserId: claims.authorizing_user_id } : {}),
|
|
identityContextId: identityRun?.activeIdentityContextId ?? null,
|
|
onBehalfOfMemberships,
|
|
source: "agent_jwt",
|
|
};
|
|
next();
|
|
return;
|
|
}
|
|
|
|
await db
|
|
.update(agentApiKeys)
|
|
.set({ lastUsedAt: new Date() })
|
|
.where(eq(agentApiKeys.id, key.id));
|
|
|
|
const agentRecord = await db
|
|
.select()
|
|
.from(agents)
|
|
.where(eq(agents.id, key.agentId))
|
|
.then((rows) => rows[0] ?? null);
|
|
|
|
if (!agentRecord || agentRecord.companyId !== key.companyId) {
|
|
next(unauthorized("Agent record is missing or belongs to another company; obtain fresh credentials and retry"));
|
|
return;
|
|
}
|
|
if (agentRecord.lifecycleHolds?.includes("setup")) {
|
|
next(unauthorized("Agent setup is incomplete and cannot authenticate"));
|
|
return;
|
|
}
|
|
if (agentRecord.status === "terminated") {
|
|
next(unauthorized("Agent is terminated and cannot authenticate"));
|
|
return;
|
|
}
|
|
if (agentRecord.status === "pending_approval") {
|
|
next(unauthorized("Agent is pending approval and cannot authenticate"));
|
|
return;
|
|
}
|
|
|
|
const responsibleUserId = normalizeOptionalString(key.responsibleUserId);
|
|
if (!responsibleUserId) {
|
|
await auditAgentKeyMissingResponsibleUser(db, {
|
|
companyId: key.companyId,
|
|
agentId: key.agentId,
|
|
keyId: key.id,
|
|
method: req.method,
|
|
url: req.originalUrl,
|
|
});
|
|
next(forbidden("Responsible user is unavailable for this agent key", {
|
|
code: "RESPONSIBLE_USER_UNAVAILABLE",
|
|
}));
|
|
return;
|
|
}
|
|
|
|
req.actor = {
|
|
type: "agent",
|
|
agentId: key.agentId,
|
|
companyId: key.companyId,
|
|
keyId: key.id,
|
|
keyScope: normalizeAgentApiKeyScope(key.scopeConfig),
|
|
onBehalfOfUserId: responsibleUserId,
|
|
onBehalfOfMemberships: await loadResponsibleUserMemberships(db, {
|
|
companyId: key.companyId,
|
|
userId: responsibleUserId,
|
|
}),
|
|
runId: runIdHeader || undefined,
|
|
source: "agent_key",
|
|
};
|
|
if (isCloudManagedInstance() && req.actor.onBehalfOfMemberships?.length) {
|
|
_res.once("finish", () => { if (_res.statusCode >= 200 && _res.statusCode < 400) idleApiActivity.record(); });
|
|
}
|
|
next();
|
|
};
|
|
return async (req, res, next) => {
|
|
// Health and task-drain requests bypass tenant admission, but resolving
|
|
// their actor can still write users, companies, memberships and key usage.
|
|
// Finish all authentication before entering the next handler: the report
|
|
// must count concurrent authentication, without counting its own auth.
|
|
const finish = beginIdleTrackedWork();
|
|
let continueRequest: (() => void) | undefined;
|
|
try {
|
|
await authenticate(req, res, (error?: unknown) => {
|
|
continueRequest = () => next(error);
|
|
});
|
|
} finally {
|
|
finish();
|
|
}
|
|
continueRequest?.();
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Minimal header accessor `resolveCloudTenantActor` needs. Express `Request`
|
|
* satisfies it directly; websocket upgrade paths adapt a raw
|
|
* `IncomingMessage` with {@link cloudActorHeaderSourceFromHeaders} since
|
|
* trusted-header authentication must work identically for upgrades — a
|
|
* cloud-proxied browser has no local Better Auth session to fall back on.
|
|
*/
|
|
export interface CloudActorHeaderSource {
|
|
header(name: string): string | undefined;
|
|
}
|
|
|
|
/** Adapts a raw header map (e.g. `IncomingMessage.headers`) to {@link CloudActorHeaderSource}. */
|
|
export function cloudActorHeaderSourceFromHeaders(
|
|
headers: Record<string, string | string[] | undefined>,
|
|
): CloudActorHeaderSource {
|
|
return {
|
|
header(name: string) {
|
|
const value = headers[name.toLowerCase()];
|
|
return Array.isArray(value) ? value[0] : value;
|
|
},
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Trusted-header actor resolution with bounded transient-connection retries.
|
|
* The tenant sync inside is idempotent end to end — every write is an
|
|
* upsert/on-conflict/delete and the write debounce records only after the
|
|
* whole sync succeeds — so replaying it after a dropped connection is safe,
|
|
* and turns a golden-path authentication 500 into a served request.
|
|
*/
|
|
export async function resolveCloudTenantActor(
|
|
db: Db,
|
|
req: CloudActorHeaderSource,
|
|
): Promise<Express.Request["actor"] | null> {
|
|
return retryIdempotentDatabaseOperation(() => resolveCloudTenantActorOnce(db, req));
|
|
}
|
|
|
|
async function resolveCloudTenantActorOnce(
|
|
db: Db,
|
|
req: CloudActorHeaderSource,
|
|
): Promise<Express.Request["actor"] | null> {
|
|
const expectedToken = process.env.PAPERCLIP_CLOUD_TENANT_SERVER_TOKEN?.trim();
|
|
if (!expectedToken) return null;
|
|
|
|
const token = req.header("x-paperclip-cloud-tenant-token")?.trim();
|
|
if (!token || !constantTimeStringEqual(token, expectedToken)) return null;
|
|
|
|
const userId = requiredCloudHeader(req, "x-paperclip-cloud-user-id");
|
|
const userEmail = requiredCloudHeader(req, "x-paperclip-cloud-user-email").toLowerCase();
|
|
const stackId = requiredCloudHeader(req, "x-paperclip-cloud-stack-id");
|
|
const stackRole = stackMembershipRole(req.header("x-paperclip-cloud-stack-role"));
|
|
const userName = req.header("x-paperclip-cloud-user-name")?.trim() || userEmail;
|
|
const paperclipCompanyId = req.header("x-paperclip-cloud-paperclip-company-id")?.trim();
|
|
const paperclipCompanyName = req
|
|
.header("x-paperclip-cloud-paperclip-company-name")
|
|
?.trim();
|
|
const companyId = cloudTenantCompanyId(stackId);
|
|
const companyName = paperclipCompanyName || humanizeCloudStackSlug(stackId);
|
|
const membershipRole = stackRole === "owner" || stackRole === "admin" ? "owner" : stackRole;
|
|
const syncFingerprint = JSON.stringify([
|
|
userEmail, userName, stackId, stackRole, paperclipCompanyId ?? "", paperclipCompanyName ?? "",
|
|
]);
|
|
let effectiveMembership: { companyId: string; membershipRole: string | null; status: string } = {
|
|
companyId,
|
|
membershipRole,
|
|
status: "active",
|
|
};
|
|
|
|
// Earlier cloud_tenant builds granted every tenant user `instance_admin`.
|
|
// Stale rows from those deployments would still elevate this user through
|
|
// the BetterAuth session path, board API keys, and the authorization
|
|
// service's own instanceUserRoles lookup — so actively purge them on every
|
|
// trusted-header authentication instead of merely no longer inserting them.
|
|
//
|
|
// Every request runs this path, static files included. The purge and the
|
|
// tenant sync stay ordered; the two reads below are independent of them, so
|
|
// they run concurrently. A loaded database then costs one round trip per
|
|
// request rather than three in a row.
|
|
const purgeAndSync = (async () => {
|
|
await db
|
|
.delete(instanceUserRoles)
|
|
.where(and(eq(instanceUserRoles.userId, userId), eq(instanceUserRoles.role, "instance_admin")));
|
|
|
|
await syncCloudTenantWrites(db, userId, syncFingerprint, async () => {
|
|
const now = new Date();
|
|
await db
|
|
.insert(authUsers)
|
|
.values({
|
|
id: userId,
|
|
name: userName,
|
|
email: userEmail,
|
|
emailVerified: true,
|
|
image: null,
|
|
createdAt: now,
|
|
updatedAt: now,
|
|
})
|
|
.onConflictDoUpdate({
|
|
target: authUsers.id,
|
|
set: { name: userName, email: userEmail, emailVerified: true, updatedAt: now },
|
|
setWhere: sql`${authUsers.name} is distinct from ${userName}
|
|
or ${authUsers.email} is distinct from ${userEmail}
|
|
or ${authUsers.emailVerified} is distinct from true`,
|
|
});
|
|
|
|
await insertCloudTenantCompany(db, { companyId, companyName, now });
|
|
if (paperclipCompanyName) {
|
|
await repairCloudTenantCompanyName(db, {
|
|
companyId,
|
|
paperclipCompanyId,
|
|
paperclipCompanyName,
|
|
now,
|
|
});
|
|
}
|
|
// Derive any legacy prefix from the repaired name, preserving the helper's
|
|
// exact legacy-marker guards and the operator's own edits.
|
|
await repairCloudTenantCompanyProvisionDefaults(db, { companyId, stackId, now });
|
|
|
|
effectiveMembership = await db.transaction(async tx => {
|
|
const [existing] = await tx.select({
|
|
status: companyMemberships.status,
|
|
membershipRole: companyMemberships.membershipRole,
|
|
}).from(companyMemberships).where(and(
|
|
eq(companyMemberships.companyId, companyId),
|
|
eq(companyMemberships.principalType, "user"),
|
|
eq(companyMemberships.principalId, userId),
|
|
)).for("update");
|
|
// Serialize authority restoration and its notice while retaining the
|
|
// guarded upsert that leaves unchanged memberships untouched.
|
|
if (!existing || existing.status !== "active" || existing.membershipRole === "viewer") {
|
|
await notifyChatCompanyAuthorityChange(tx, { companyId });
|
|
await notifyMuseContactVerification(tx, { companyId, operatorId: userId });
|
|
}
|
|
return tx
|
|
.insert(companyMemberships)
|
|
.values({ companyId, principalType: "user", principalId: userId, status: "active", membershipRole, updatedAt: now })
|
|
.onConflictDoUpdate({
|
|
target: [companyMemberships.companyId, companyMemberships.principalType, companyMemberships.principalId],
|
|
set: { status: "active", membershipRole, updatedAt: now },
|
|
setWhere: sql`${companyMemberships.status} is distinct from 'active'
|
|
or ${companyMemberships.membershipRole} is distinct from ${membershipRole}`,
|
|
})
|
|
.returning({ companyId: companyMemberships.companyId, membershipRole: companyMemberships.membershipRole, status: companyMemberships.status })
|
|
// PostgreSQL returns no row for an unchanged ON CONFLICT ... WHERE.
|
|
// That condition itself proves the existing role/status match the input.
|
|
.then((rows) => rows[0] ?? { companyId, membershipRole, status: "active" });
|
|
});
|
|
|
|
// An unchanged membership can still lack default grants after a partial
|
|
// sync or an older deployment; keep this repair inside every actual sync.
|
|
await ensureHumanRoleDefaultGrants(db, {
|
|
companyId,
|
|
principalId: userId,
|
|
membershipRole: effectiveMembership.membershipRole ?? membershipRole,
|
|
grantedByUserId: null,
|
|
});
|
|
});
|
|
})();
|
|
|
|
// The stack's seeded company is only where Cloud provisioned this user.
|
|
// Companies created afterwards on the instance (imports, in-app company
|
|
// creation) attach real membership rows for the user, so union those with
|
|
// the pinned primary — the same active-membership scope a locally
|
|
// authenticated session actor carries. Strictly this user's own rows; the
|
|
// membership-creating flows seed their own permission grants, so nothing
|
|
// needs seeding per request here. The sync above writes only the primary
|
|
// company's membership, which this read excludes, so it need not wait for
|
|
// the sync. A read failure degrades to the pinned primary instead of
|
|
// blocking authentication, mirroring the fail-closed owner-elevation
|
|
// resolution below.
|
|
const additionalMembershipsRead: Promise<{ companyId: string; membershipRole: string | null; status: string }[]> =
|
|
loadActiveUserCompanyMemberships(db, userId).then(
|
|
(rows) => rows.filter((row) => row.companyId !== companyId),
|
|
(err) => {
|
|
logger.warn(
|
|
{ err, userId, stackId },
|
|
"Failed to load cloud tenant user's company memberships; scoping actor to the stack's primary company",
|
|
);
|
|
return [];
|
|
},
|
|
);
|
|
// Computed per request, never persisted: the stack owner is elevated to
|
|
// instance admin of their own dedicated instance only while the
|
|
// `enableOwnerInstanceAdmin` flag is on. Non-owner stack roles stay
|
|
// company-scoped. Turning the flag off de-elevates on the next request —
|
|
// there is no role row to clean up.
|
|
const ownerInstanceAdminRead = resolveOwnerInstanceAdmin(db, stackRole);
|
|
|
|
// Settle every started statement before an error leaves: the caller stops
|
|
// counting authentication work for idle and drain checks when this throws,
|
|
// and the owner-settings read can insert the missing settings row.
|
|
const [synced, membershipsRead, ownerRead] = await Promise.allSettled([
|
|
purgeAndSync,
|
|
additionalMembershipsRead,
|
|
ownerInstanceAdminRead,
|
|
]);
|
|
if (synced.status === "rejected") throw synced.reason;
|
|
if (membershipsRead.status === "rejected") throw membershipsRead.reason;
|
|
if (ownerRead.status === "rejected") throw ownerRead.reason;
|
|
const additionalMemberships = membershipsRead.value;
|
|
const isInstanceAdmin = ownerRead.value;
|
|
|
|
return {
|
|
type: "board",
|
|
userId,
|
|
userName,
|
|
userEmail,
|
|
companyIds: [companyId, ...additionalMemberships.map((row) => row.companyId)],
|
|
memberships: [
|
|
{
|
|
companyId,
|
|
membershipRole: effectiveMembership.membershipRole ?? membershipRole,
|
|
status: effectiveMembership.status,
|
|
},
|
|
...additionalMemberships,
|
|
],
|
|
isInstanceAdmin,
|
|
canManageExperimentalSettings: stackRole === "admin",
|
|
source: "cloud_tenant",
|
|
};
|
|
}
|
|
|
|
function requiredCloudHeader(req: CloudActorHeaderSource, name: string): string {
|
|
const value = req.header(name)?.trim();
|
|
if (!value) {
|
|
throw new Error(`Missing trusted Cloud tenant header ${name}`);
|
|
}
|
|
return value;
|
|
}
|
|
|
|
function stackMembershipRole(value: string | undefined): "owner" | "admin" | "member" | "support" {
|
|
if (value === "owner" || value === "admin" || value === "member" || value === "support") {
|
|
return value;
|
|
}
|
|
throw new Error("Invalid trusted Cloud tenant stack role");
|
|
}
|
|
|
|
function constantTimeStringEqual(left: string, right: string): boolean {
|
|
const leftBuffer = Buffer.from(left);
|
|
const rightBuffer = Buffer.from(right);
|
|
return leftBuffer.length === rightBuffer.length && timingSafeEqual(leftBuffer, rightBuffer);
|
|
}
|
|
|
|
function cloudTenantCompanyId(stackId: string): string {
|
|
return cloudTenantPrimaryCompanyId(stackId);
|
|
}
|
|
|
|
export function humanizeCloudStackSlug(stackId: string): string {
|
|
const slug = stackId
|
|
.trim()
|
|
.replace(/^paperclip-stack-/i, "")
|
|
.replace(/^stack-/i, "");
|
|
const displayName = slug
|
|
.split(/[-_]+/)
|
|
.filter(Boolean)
|
|
.map((part) => `${part[0]?.toUpperCase() ?? ""}${part.slice(1)}`)
|
|
.join(" ");
|
|
return displayName || "Workspace";
|
|
}
|
|
|
|
export function isKnownBadCloudCompanyName(
|
|
name: string,
|
|
ids: { companyId: string; paperclipCompanyId?: string },
|
|
): boolean {
|
|
const normalized = name.trim();
|
|
return (
|
|
/^paperclip-stack-.+/i.test(normalized) ||
|
|
/^stack-.+\s+paperclip$/i.test(normalized) ||
|
|
normalized === ids.companyId ||
|
|
(ids.paperclipCompanyId !== undefined &&
|
|
normalized === ids.paperclipCompanyId)
|
|
);
|
|
}
|
|
|
|
async function repairCloudTenantCompanyName(
|
|
db: Db,
|
|
input: {
|
|
companyId: string;
|
|
paperclipCompanyId?: string;
|
|
paperclipCompanyName: string;
|
|
now: Date;
|
|
},
|
|
): Promise<void> {
|
|
try {
|
|
const existing = await db
|
|
.select({ name: companies.name })
|
|
.from(companies)
|
|
.where(eq(companies.id, input.companyId))
|
|
.then((rows) => rows[0]);
|
|
if (
|
|
!existing ||
|
|
!isKnownBadCloudCompanyName(existing.name, {
|
|
companyId: input.companyId,
|
|
paperclipCompanyId: input.paperclipCompanyId,
|
|
})
|
|
) {
|
|
return;
|
|
}
|
|
await db.transaction(async (tx) => {
|
|
const [updated] = await tx
|
|
.update(companies)
|
|
.set({ name: input.paperclipCompanyName, updatedAt: input.now })
|
|
.where(
|
|
and(
|
|
eq(companies.id, input.companyId),
|
|
// A user may rename the company between the read above and this
|
|
// repair. Match the exact observed machine name so that concurrent
|
|
// genuine renames always win.
|
|
eq(companies.name, existing.name),
|
|
),
|
|
)
|
|
.returning({ id: companies.id });
|
|
if (!updated) return;
|
|
|
|
await tx.insert(activityLog).values({
|
|
companyId: input.companyId,
|
|
actorType: "system",
|
|
actorId: "cloud-tenant-auth",
|
|
action: "company.updated",
|
|
entityType: "company",
|
|
entityId: input.companyId,
|
|
details: {
|
|
source: "cloud_tenant_auth",
|
|
reason: "legacy_machine_name_repair",
|
|
previousName: existing.name,
|
|
name: input.paperclipCompanyName,
|
|
},
|
|
});
|
|
});
|
|
} catch (err) {
|
|
logger.warn(
|
|
{ err, companyId: input.companyId },
|
|
"Failed to repair legacy Cloud tenant company name",
|
|
);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Claims the tenant company row for this stack.
|
|
*
|
|
* The prefix derives from the company name, exactly as it does for a
|
|
* self-hosted company. Each attempt is a standalone INSERT, so a failed
|
|
* attempt is its own implicit transaction and cannot poison a surrounding
|
|
* one. `onConflictDoNothing` only absorbs the `companies.id` conflict — a
|
|
* prefix that another company already holds still raises `23505`, so the loop
|
|
* moves on to the next suffix.
|
|
*/
|
|
async function insertCloudTenantCompany(
|
|
db: Db,
|
|
input: { companyId: string; companyName: string; now: Date },
|
|
): Promise<void> {
|
|
const base = deriveIssuePrefixBase(input.companyName);
|
|
for (let attempt = 1; attempt <= MAX_ISSUE_PREFIX_ATTEMPTS; attempt += 1) {
|
|
try {
|
|
await db
|
|
.insert(companies)
|
|
.values({
|
|
id: input.companyId,
|
|
name: input.companyName,
|
|
description: null,
|
|
status: "active",
|
|
issuePrefix: `${base}${issuePrefixSuffixForAttempt(attempt)}`,
|
|
updatedAt: input.now,
|
|
})
|
|
.onConflictDoNothing({
|
|
target: companies.id,
|
|
});
|
|
return;
|
|
} catch (error) {
|
|
if (!isIssuePrefixConflict(error)) throw error;
|
|
}
|
|
}
|
|
throw new Error("Unable to allocate a unique issue prefix for the tenant company");
|
|
}
|
|
|
|
/**
|
|
* The issue prefix that pre-name-derivation builds gave a tenant company.
|
|
*
|
|
* This derivation survives only as the detector for the one-time repair
|
|
* below. Nothing mints a prefix this way any more.
|
|
*/
|
|
function legacyProvisionedIssuePrefix(stackId: string): string {
|
|
const hash = createHash("sha256").update(stackId).digest("hex").slice(0, 4).toUpperCase();
|
|
return `PC${hash}`;
|
|
}
|
|
|
|
/** The placeholder description that pre-name-derivation builds wrote. */
|
|
const LEGACY_PROVISIONED_DESCRIPTION_PREFIX = "Provisioned by Paperclip Cloud for stack ";
|
|
|
|
/**
|
|
* One-time repair for companies claimed by a pre-name-derivation build.
|
|
*
|
|
* Those companies carry an opaque hash prefix and a placeholder description
|
|
* that the operator never chose. Re-derive the prefix from the company's
|
|
* current name, re-key the stored issue and case identifiers onto it, and drop
|
|
* the placeholder. Both guards stop matching once the repair lands, so a later
|
|
* pass is a no-op.
|
|
*/
|
|
async function repairCloudTenantCompanyProvisionDefaults(
|
|
db: Db,
|
|
input: { companyId: string; stackId: string; now: Date },
|
|
): Promise<void> {
|
|
try {
|
|
const existing = await db
|
|
.select({
|
|
name: companies.name,
|
|
issuePrefix: companies.issuePrefix,
|
|
description: companies.description,
|
|
})
|
|
.from(companies)
|
|
.where(eq(companies.id, input.companyId))
|
|
.then((rows) => rows[0]);
|
|
if (!existing) return;
|
|
|
|
const legacyPrefix = legacyProvisionedIssuePrefix(input.stackId);
|
|
const legacyDescription = existing.description?.startsWith(LEGACY_PROVISIONED_DESCRIPTION_PREFIX)
|
|
? existing.description
|
|
: null;
|
|
if (existing.issuePrefix !== legacyPrefix) {
|
|
if (!legacyDescription) return;
|
|
// The prefix was already re-derived, so only the placeholder is left.
|
|
await db
|
|
.update(companies)
|
|
.set({ description: null, updatedAt: input.now })
|
|
.where(and(
|
|
eq(companies.id, input.companyId),
|
|
eq(companies.description, legacyDescription),
|
|
));
|
|
return;
|
|
}
|
|
|
|
await db.transaction(async (tx) => {
|
|
const candidate = await pickAvailableIssuePrefix(tx, deriveIssuePrefixBase(existing.name));
|
|
if (!candidate || candidate === legacyPrefix) return;
|
|
|
|
const [updated] = await tx
|
|
.update(companies)
|
|
.set({
|
|
issuePrefix: candidate,
|
|
...(legacyDescription ? { description: null } : {}),
|
|
updatedAt: input.now,
|
|
})
|
|
.where(and(
|
|
eq(companies.id, input.companyId),
|
|
// A user may rename the company between the read above and this
|
|
// repair. Match the exact observed legacy prefix so that concurrent
|
|
// genuine renames always win.
|
|
eq(companies.issuePrefix, legacyPrefix),
|
|
))
|
|
.returning({ id: companies.id });
|
|
if (!updated) return;
|
|
|
|
const rekeyed = await rekeyCompanyIssueIdentifiers(tx, {
|
|
companyId: input.companyId,
|
|
fromPrefix: legacyPrefix,
|
|
toPrefix: candidate,
|
|
});
|
|
|
|
await tx.insert(activityLog).values({
|
|
companyId: input.companyId,
|
|
actorType: "system",
|
|
actorId: "cloud-tenant-auth",
|
|
action: "company.updated",
|
|
entityType: "company",
|
|
entityId: input.companyId,
|
|
details: {
|
|
source: "cloud_tenant_auth",
|
|
reason: "legacy_provision_defaults_repair",
|
|
previousIssuePrefix: legacyPrefix,
|
|
issuePrefix: candidate,
|
|
descriptionCleared: legacyDescription !== null,
|
|
issuesRekeyed: rekeyed.issues,
|
|
casesRekeyed: rekeyed.cases,
|
|
},
|
|
});
|
|
});
|
|
} catch (err) {
|
|
logger.warn(
|
|
{ err, companyId: input.companyId },
|
|
"Failed to repair legacy tenant company provisioning defaults",
|
|
);
|
|
}
|
|
}
|
|
|
|
export function requireBoard(req: Express.Request) {
|
|
return req.actor.type === "board";
|
|
}
|