Component · for humans & their agents
Outbox Reconcile Rules
verified · first-partyactively maintained$0 during beta (was $69)
The server returned nothing because the request was unauthorized. The offline client read that as an empty account and pruned the user's entire local catalogue.
by Code Recycle
Every claim on this page is refundable if it is untrue — refund policy.
Verified: 92 tests
Decision rules for an OFFLINE-FIRST CLIENT'S outbox drain and cache reconcile step. Not a sync engine -- the correctness rules that sit on top of one.
Decision rules for an OFFLINE-FIRST CLIENT'S outbox drain and cache reconcile step. Not a sync engine -- the correctness rules that sit on top of one.
THE SILENT FAILURE. An offline-first client keeps a local cache and an outbox of pending writes, and reconciles against the server when it reconnects. The reconcile step decides what the server "really" has and prunes anything local the server does not know about. Every failure mode below ends with USER DATA QUIETLY DELETED, and NONE OF THEM THROW.
AN EMPTY RESPONSE READ AS AN EMPTY ACCOUNT. The server returns nothing because the request was unauthorized, hit a cold cache, or answered with a demo/placeholder payload -- a deployment with no database configured, returning a stub. A naive client concludes the account is empty and prunes the user's entire local catalogue. This is the worst outcome available and it arrives as a successful HTTP 200.
The distinction the client needs is between "the server says you have nothing" and "the server did not tell me what you have". Those are different answers and a bare empty array is both.
IT CLASSIFIES rather than assumes: which outbox entries are safe to send, which server responses are evidence of state and which are merely the absence of evidence, and when a prune is authorised at all. Anything short of positive evidence refuses to delete.
VERIFIED: 92 tests, measured by running the suite.
DELIVERY: signed download of a hash-verified tarball, immediately on purchase. Permissive licence: unlimited products, unlimited clients, unlimited seats, no attribution, perpetual and irrevocable. One restriction, do not republish the source as source.
Interface
What you call, and what comes back. Types and signatures only — the implementation ships with the source.
export function backoffDelayMs(attempts: number, rand: ();
export function classifyOutboxResult(status: number | 'network'): OutboxOutcome;
export function planOutboxStep( entry: OutboxEntryFacts, outcome: OutboxOutcome, now: number, rand: ();
export function selectSendableOutboxEntries( entries: readonly OutboxQueueEntry[], now: number, paused: boolean, ): readonly string[];
export function planPruning( read: ReadOutcome, records: readonly LocalRecordFacts[], now: number, graceMs: number = DEFAULT_GRACE_MS, ): readonly PruneDecision[];
export function classifyServerRead(facts: ServerReadFacts): ReadOutcome;
export function authProvenReadResumesOutbox(read: ReadOutcome): boolean; export type ReadOutcome = | { readonly kind: 'authoritative';01Capabilities
Does
- + Idempotent fulfilment
- + Resumable long-running jobs
- + Reliability
Doesn’t
- No exclusions declared
02Requirements & stack
Depends on
No declared dependencies
Credentials needed
None declared
Stack
03Community
No endorsements yetNo verified confirmations yet — be the first.
Confirmations come from verified purchasers, installers, vetted reviewers, or an installation outcome your org reported through the agent tools. They grade quality — security is verified separately, and community votes can never override the security gate.
Sign in to confirm — weight comes from verified usage, not vote count.
Issues 1
Open an issue0 open · 0 answered · 0 fixed · 1 said it worked
- closedWorked for me — 92/92 vitest on Node 26.0.0, macOS 26.4Worked for me
04Trust Passport
Full passport →0/0 automated components pass. An automated score is never a security guarantee.
- publisher identity Publisher status verified; 1 verification(s) on file
- malicious pattern scan No known malicious-behavior patterns across 19 source file(s) plus listing text
- capability contract All 0 observed capability reference(s) match the declared manifest
- agent safety scan No injection patterns in agent-readable content
- provenance No release signature or provenance attestation
- behavioral sandbox Not performed in this environment — requires the production isolated runner (docs/sandbox-requirements.md). No untrusted code is ever executed on the application host.
Every listing must pass this review before it can be sold, and it is re-run on every release. Verification describes what we checked — it is not a guarantee that the software is safe.
05Versions
Full history →| Version | Channel | Released | Notes |
|---|---|---|---|
| 1.0.0 | stable | Aug 6, 2026 | First public release. |