Home / Studio notes / Workspace guide
Guide • Updated July 28, 2026

Organize RDM around decisions people already make.

Good architecture reduces hesitation: operators recognize where to look, which entry to trust, and how to proceed.

1. Observe retrieval behavior.

Ask operators to find five common and three unusual targets. Note the words, filters, locations, and cues they use.

  • Known target by name
  • Unknown target by service
  • Environment-specific session
  • Shared tool or website
  • Recently changed system

2. Separate stable boundaries from attributes.

Vault

Audience, owner, risk, or lifecycle boundary.

Folder

Predictable navigation and inherited context.

Tag

Cross-cutting attributes and alternate retrieval.

Field

Structured identity, owner, and environment detail.

3. Write a naming grammar.

A grammar describes the information a name carries, its order, abbreviations, case, and exceptions. Test it on real RDP, SSH, web, VPN, and administrative entries.

Weak nameQuestion it leavesBetter context
Server 12Which service and environment?Purpose + environment + location
Admin PortalFor which tenant or platform?System + scope + role
VPN NewNew compared with what?Destination + audience + owner

4. Design the lifecycle.

Define who can create entries, who validates them, how ownership changes, what review proves, and how retirement handles dependencies.

  1. Request or identify need
  2. Create from approved template
  3. Validate with representative role
  4. Review on schedule or change
  5. Retire with owner confirmation

Common mistakes

Copying the organization chart

Teams change faster than many systems. Prefer stable services, ownership, audience, or operational boundaries.

Using folders for every attribute

Deep nesting makes alternate search paths difficult. Use tags and structured fields for cross-cutting properties.

Leaving exceptions anonymous

Every exception needs a reason, owner, approval, and review date.

Turn the guide into your workspace standard.

Mesa can facilitate the decisions and prototype the result.

Start the blueprint