How agent-ready is 1Password?

1Password's overall agentability score of 43/100 indicates that autonomous AI agents will encounter moderate to significant friction when attempting to parse documentation, navigate product information, and execute tasks on the public website. The homepage scores slightly better at 43, pricing at 47, and documentation at 34, suggesting inconsistent agent-readiness across key user journeys.

The audit identifies transparency (18/100) and shadow UI avoidance (15/100) as critical weaknesses, meaning agents struggle to understand the provenance of information and encounter UI patterns that are difficult to programmatically detect. Defaults (27/100) and chunking (41/100) also score poorly, indicating content is not structured to clearly signal authentication requirements or present information in discrete, answerable units. Machine readability (78/100) and status communication (65/100) represent relative strengths.

These scores suggest that while 1Password's site is technically parseable, agents will frequently fail to determine which features require sign-in, extract structured answers from documentation, or trace the source of displayed information—limiting autonomous task completion and requiring more human intervention than a highly agent-ready platform would need.

To improve agentability, 1Password should focus on the following actionable changes:

• Clearly distinguish what features and content are usable without authentication from what requires sign-in, preventing agents from attempting unauthorized actions or misrepresenting capabilities to users.
• Restructure documentation headings as direct questions they answer (e.g., 'How do I cancel my subscription?') to enable agents to quickly match user intent with relevant content.
• Lead each documentation section with the core answer in the first one to two sentences before elaborating, allowing agents to extract key information without parsing entire articles.
• Offer expandable drill-downs for important results—provide a one-line summary that agents can surface immediately, with optional detailed expansions for context.
• Provide labeled pause, cancel, or stop controls for long-running actions such as imports or exports, using accessible names like 'Cancel import' rather than unlabeled icons that agents cannot reliably identify.
• Attach source references, input citations, or identifiers to generated or dynamic outputs so agents can trace provenance and assess information reliability.