Skip to content
Harborby Signalry
CA
Agent

Who, after the job

Clerkshift Agent

How and who for the hotel job—verified when attested.

Reservation modificationsA narrow agent that proposes date, rate, and add-on changes — a human posts anything that touches the folio.
8.0

Post-job

9 reviews

Not certifiedA catalog listing is not evidence. Harbor by Signalry has not checked this provider. Evidence tiers

Leave a record

The next hotel starts from what you leave

Harbor by Signalry stays hotels only. A description, a claim, and a reported result each leave a stable id behind. Nothing here is a price, a customer, or a verification.

  • Buyer

    Describe your job

    No account. Harbor by Signalry opens the closest catalog brief, when not to buy, and a shortlist. A weak sentence stays on the catalog.

    Open a brief
  • Supplier

    Claim this listing

    Company, jobs served, regions, and a hotel reference. The packet is queued as unverified pending. The form cannot mark it verified.

    Claim this listing
  • Hotel

    Record a result

    A metric, before and after, and the dates. It is shown as reported until a person checks the evidence. Verified Outcome is not insurance.

    Record a result

HQ

Remote · agent

Coverage

US independents with an API-capable PMS

Typical fee

$0.40–0.90 / handled request + monthly floor

PMS shape (demo)

opera, mews

Keys

under-80, 80-200, 200-plus

Flag

independent, soft-brand, branded

About

Clerkshift is an AI agent, not a chatbot. It drafts reservation modifications against PMS and CRS rules, then queues them for a named clerk. Harbor scores it on error rate and 'did a human still own the folio,' not on deflection.

AdviceDesks drowning in date-change email.

ClaimedRivermark: 73% of mods drafted by the agent, 100% human-posted in month one.

Best for

Desks drowning in date-change email.

Not for

Hotels that want an unmanned reservations office.

Score breakdown

Outcome · 50%

8.1

Integrity · 30%

8.3

Fit · 20%

7.6

Trust & methodology — post-job, no paid rank.

Evidence

  • · Rivermark: 73% of mods drafted by the agent, 100% human-posted in month one.
  • · A silent auto-post is a scoring fail.

Tradeoffs they name

  • · An agent will not staff a sold-out Saturday.
  • · Needs a clean reservations inbox.

Jobs they appear on

Peer cases