Is a feature flag a kill switch?

It can be, if every trigger, worker and retry checks it before doing work, and if the review proves they do. A flag that only the web app reads is not a kill switch for a background worker.

How fast should it be?

Fast enough that the spend and damage during the stop are acceptable to you. Time it on staging and write the number down; that measured time is what the report records.

Question: kill switch

What should an AI agent kill switch actually do?

In a Mandiant case study, an accounting agent made more than 15,000 high-cost API calls in under an hour, about $50,000 in cloud charges (Mandiant AI Risk and Resilience report, as reported by Help Net Security on 2026-09-16 (helpnetsecurity.com/2026/09/16/google-mandiant-enterprise-ai-security-risks-report)). At that rate, the minutes it takes to find the off switch are the cost.

Request a Pre-Flight Review See the six guardrails

Last updated 2026-09-26 · Agent Guardrail Desk, Denver, Colorado

Short answer

It should stop every run in flight, block new runs, cancel work already queued on the server, and be held by a named person who can use it from a runbook in minutes. Closing the laptop or killing the client process often does not stop server-side work.

The four things it must stop

Runs in flight: the current step finishes or aborts, and no next step starts.
New runs: triggers, schedules and webhooks are refused while the switch is on.
Queued work: jobs already handed to a queue, a worker or a hosted agent service.
Retries: automatic retry logic, which can restart a run you just stopped.

Who holds it

One named person per shift, written into the runbook, with a second named person as backup. The switch should not need the engineer who built the agent.

How to test it

Pull the switch during a live run on a staging copy and time it. Pass: the run in flight stops, no new run starts, queued work is cancelled, and the time to stop is written down. This is the Kill Switch acceptance test in the Six-Guardrail Pre-Flight.

Want it tested for you?

This test is one of six in the Six-Guardrail Pre-Flight. $1,500 per agent or automation, once. Delivered in 5 business days after read-only access to a staging copy.

See the Pre-Flight Review scope and price

Questions

Questions about this

Is a feature flag a kill switch?: It can be, if every trigger, worker and retry checks it before doing work, and if the review proves they do. A flag that only the web app reads is not a kill switch for a background worker.
How fast should it be?: Fast enough that the spend and damage during the stop are acceptable to you. Time it on staging and write the number down; that measured time is what the report records.

Every page links back to the Agent Guardrail Desk home page, where the price, the rulebook and the request form live.

The review, per situation

SituationPre-Flight for Agent BuildersEngineers shipping tool-using agents, multi-agent pipelines and MCP servers. The loop stops at the cap; the hidden instruction is refused or held for approval.SituationPre-Flight for AI-Built AppsFounders and makers shipping apps built with Lovable, Bolt, Cursor or Claude Code on Supabase. The request is refused; no service key ships to the browser.SituationPre-Flight for No-Code AutomationsOps owners running Zapier, Make or n8n flows from forms and webhooks into a CRM. The run stops and alerts instead of writing an empty record.SituationPre-Flight for Platform TeamsPlatform, security and ops leads whose internal agents can write to CRM, ticketing and docs. The write is refused; the agent's role grants only what its job needs.

Buyer questions, answered

QuestionHow do I cap what an AI agent can spend?Where a spend cap has to live, what it counts, and the test that proves it stops a loop.QuestionHow do I catch a renamed form field before it writes empty records?Silent breaks in automations: what to check before each write, and the test that proves it.QuestionHow do I check an AI-built app is secure before real users?The pre-launch routine for Lovable, Bolt, Cursor and Claude Code apps, and what to test after it.

Price and sample

Start hereThe Pre-Flight Review: price, scope and timeline$1,500 per agent, report in 5 business days, credited in full against a Fix Sprint within 30 days.Start hereSee a sample Pre-Flight ReportThe report format, filled in for a placeholder agent. Demonstration, not a client result.HomeAgent Guardrail Desk: the Six-Guardrail Pre-FlightThe whole offer on one page: the six guardrails, the price, the rulebook and the form.

Next step

Request a Pre-Flight Review for one agent

$1,500 per agent or automation, once. Delivered in 5 business days after read-only access to a staging copy. If you book a Fix Sprint within 30 days, the $1,500 is credited in full against it.

Request a Pre-Flight Review See a sample report