How agent-ready is LastPass?
Independent agentability audit of LastPass, scored across the 8 principles of Agent Factors Engineering — how well AI agents can parse, navigate, and operate it.
Audit summary
LastPass received an overall agentability score of 37/100 on its homepage, with similar scores across pricing (42/100) and documentation (34/100). These scores indicate that autonomous AI agents will encounter significant friction when attempting to parse content, navigate flows, and complete tasks on the platform.
The audit identified two critical gaps: shadow UI avoidance and transparency both scored 0/100, meaning the site provides no machine-readable activity logs, confidence indicators, or explanations for system actions. Machine readability performed best at 72/100, though opportunities remain to adopt semantic HTML landmarks. Status updates (35/100), chunking (41/100), control mechanisms (47/100), defaults (50/100), and clean handoffs (50/100) all fall in the mid-range, pointing to inconsistent implementation of agent-friendly patterns.
These scores suggest that while basic content extraction may succeed, agents will struggle with dynamic interactions, understanding system state, and reliably completing multi-step workflows without human intervention.
Score by principle
Key findings
How LastPass could improve its score
LastPass can improve agentability by addressing the following issues identified in the audit:
- Replace generic
<div>and<span>elements with semantic HTML5 landmarks such as<header>,<nav>,<main>,<article>,<section>, and<footer>to improve machine readability. - Wrap status updates and notifications in ARIA live regions using
aria-live,role="status", orrole="alert"so agents can detect state changes programmatically. - Expose an activity log or audit trail in machine-readable format (such as a JSON event log alongside a visible activity feed) that records system actions and provides a 'why this result' affordance for agents to trace decision paths.
- Add confidence indicators or certainty metadata to key outputs, such as a 'verified vs. best-guess' label in the UI or a confidence field in API responses, to help agents assess data reliability.
- Provide labeled pause, cancel, or stop controls for long-running operations with accessible names like 'Cancel import' rather than unlabeled icons, ensuring agents can programmatically interrupt processes.
- Rephrase section headings as direct questions they answer (for example, 'How do I cancel my subscription?') to improve content chunking and help agents quickly locate relevant information.
Work at LastPass? Re-audit any page free.
Scores refresh automatically when we re-crawl — or run an instant audit on any URL now.
Run a free auditEmbed this score
Show your agent-readiness in your docs or README. The badge links back to this live report.