remora/docs/api

Remora API implementation contract

The HTTP and WebSocket artifacts in this directory are generated from contracts/typespec/main.tsp. Edit and regenerate that TypeSpec source instead of editing these files. The generated artifacts describe the current wire contract but are not authoritative over Remora.tla or architecture.md. Conflicts are corrected here.

Run npm ci --prefix contracts/typespec and npm run generate --prefix contracts/typespec after changing the contract. CI uses the corresponding check command to reject stale artifacts.

OpenAPI 3.2.0 describes HTTP operations and JSON Schema Draft 2020-12 describes WebSocket frames. The API contract version is 1.0.0; additive fields require clients to tolerate unknown response fields only after a contract version that permits them, while incompatible changes use a new major path or media type.

Idempotency

Every public domain mutation requires Idempotency-Key. The identity is caller subject, operation, and key. The server canonicalizes the semantic request, including the authorization target, with RFC 8785 and stores its SHA-256 fingerprint and complete response atomically with the mutation. The default retention is 24 hours and is never shorter than the advertised Idempotency-Expires-At.

Event-ticket issuance is an ephemeral authentication exchange rather than a domain mutation. Each retry independently mints a short-lived credential; it does not refine a JobProtocol effect or enter the public mutation ledger.

The same identity and fingerprint replays the original status, resource, and body after duplicate or ambiguous delivery. A different fingerprint returns idempotency_conflict and cannot affect either request. Naturally idempotent deletion and cancellation semantics still use this request identity.

Internal claim uses job ID plus operation_id because no attempt exists yet. Renew, progress, and outcome use attempt ID, fencing token, worker incarnation, and operation_id. Each operation identity retains a request fingerprint. An exact replay has the original effect; a changed worker, fence, or payload returns precondition_failed without mutation. First observation is accepted only when the corresponding JobProtocol action is enabled. Later eligibility changes do not alter a stored response, while a fresh operation is evaluated against current eligibility.

Renewal follows RenewLease: it extends an existing live coordination lease only while the attempt remains current durable authority, AttemptExecuting, and non-terminal. It never creates or restores a claim. After authority loss the lease ages out or is released; it cannot be extended. Progress and outcome operations require current fenced authority and return precondition_failed for stale or non-authoritative tuples.

Activation returns the live lease expiry and a recommended renewal time. Renewal returns the corresponding updated projection. These timestamps schedule worker behavior; the coordination store remains authoritative for whether the lease is live.

A successful claim returns the attempt authority and the attempt's ExecutionEnvelope. The worker verifies that their job, attempt, fence, and worker incarnation agree before source or target I/O. Resolved-source and checkpoint descriptors identify durable streamable state; they are not reconstructed from coordination data.

Resolved-source and checkpoint references carry their descriptor kind, schema version, byte length, and SHA-256. The service accepts them only when their kind, ownership, integrity, and fenced attempt authority match the mutation.

RemoraApiMutation models this durable operation ledger around an abstract successful JobProtocol effect. It does not redefine job phases or attempt-authority transitions. Its Rejected response abstracts the endpoint-specific idempotency_conflict and precondition_failed codes enforced by the executable contract.

Authorization and errors

Public operations apply the authorization rules in the architecture. List is filtered before pagination. A missing and an unauthorized resource both return not_found; callers cannot probe job existence. Cursors and event replay are bound to the authenticated subject and filter.

ApiError fields are an allowlist. Messages and field diagnostics never carry credential values, bearer material, URL userinfo, raw upstream response bodies, internal stack traces, or unredacted configured endpoints.

Credential creation uses a closed material shape selected by kind; rotation must use the existing kind's shape. Custom HTTP headers use lowercase canonical names and reject connection-nominated, routing, framing, cookie, proxy, and authentication headers.

Jobs, pages, and results

Job.phase uses the exact JobProtocol phase vocabulary; this contract does not define phase transitions. lifecycle, current authoritative attempt, current matching progress, reusable-checkpoint projection, result, and failure are separate fields. Progress is present only when its attempt ID and fence match the authoritative attempt in the same projection.

Job pages order by (created_at DESC, id DESC). The first request fixes a membership snapshot, authorization/filter fingerprint, and high-water mark. next_cursor is an opaque integrity-protected keyset cursor for that snapshot; new jobs do not enter later pages. Invalid, altered, cross-subject, or cross-filter cursors return the corresponding structured cursor error.

A result endpoint returns only a visible CommittedResult. Tombstoned, unauthorized, absent, and never-committed results are not exposed. The hard delete endpoint returns the durable deletion resource established by the architecture.

Exact member access returns a non-cacheable time-limited URL after current authorization is revalidated. This JSON exchange lets a bearer-authenticated browser deliberately navigate to the target URL without relying on an unobservable cross-origin fetch redirect.

Before hard deletion, job reads return the complete job projection. Once only the permanent minimal record remains, authorized reads and deleted-job pages use the distinct tombstone projection rather than inventing discarded job fields.

A deletion whose exact remaining work cannot progress because target access or configuration is unavailable enters blocked with a redacted problem. It retains the deletion obligation and can return to active processing when the dependency is restored. Owner access remains authorized through the concrete target retained in the tombstone.

Worker failure outcomes accept only attempt-reportable runtime classes and do not accept a caller-supplied retryable flag. The service derives the durable disposition from the class mapping in the architecture. Validation, cancellation, and lost-attempt projection use their dedicated boundaries.

WebSocket replay

An HTTP job or list snapshot carries event_cursor at its consistency watermark. The client connects to /v1/events with that cursor; there is no subscription frame. Server frames are ordered by increasing sequence for a connection and carry an opaque next cursor. Delivery is at least once, and clients deduplicate event IDs. Each connection starts at sequence one and increments it by one for every frame, including heartbeats and replayed event IDs; a client resnapshots when it observes a gap.

Browser clients first create a short-lived, single-use event ticket through an OIDC bearer-authenticated HTTP request. The server stores only its digest and binds it to the authenticated principal and exact allowed origin. Redemption atomically consumes it during the WebSocket handshake. Native clients may authenticate that handshake directly with the OIDC bearer header.

job_changed is an invalidation with an attempt-aware display hint, not a resource replacement or patch. A client compares job_revision with the last HTTP snapshot revision and fetches a new authorized snapshot when it is newer. It advances its applied revision only from that snapshot. Matching progress may be displayed provisionally while the fetch is pending.

deletion_changed is likewise an invalidation. Its state is a display hint; the deletion resource remains the authoritative HTTP projection.

Replay applies current authorization before delivery. If a cursor is absent, invalid, expired, authorization-incompatible, or no longer available after coordination loss, the server sends resync_required; the client obtains a new HTTP snapshot and reconnects from its cursor. Progress frames for a stale attempt/fence are suppressed before emission and ignored defensively by the client.