If you've read our overview of the Analyze Site and Buildable Envelope agents, you already know what they do. This post skips the explanation and goes straight to how they behave on real project types, with the kind of variety that makes site conditions interesting and occasionally frustrating.

Four typologies, four different sets of zoning conditions. Here's what changes (and what stays the same) when you run these agents across them.

How do site agents handle a dense urban infill site?

Urban lots stress-test these agents the most. The site is small, multiple zoning overlays stack on top of each other, and a heritage or character precinct often adds restrictions that don't show up cleanly in the base zoning.

The site: A 900 sqm corner lot in a dense city centre. Base zoning permits mixed-use development. A heritage precinct overlay applies, adding a facade-line setback on the street-facing edges. The site sits on a transit corridor that brings an FAR bonus.

What you import: The site boundary and parcel data, loaded via Load Site. The zoning data that comes with the site captures the base zoning, the overlay, and the corridor bonus.

What Analyze Site surfaces: Setbacks for each boundary, including the additional heritage facade setback. The FAR figure including the corridor bonus, alongside the base figure. The height limit. A summary noting the heritage overlay as a constraint to plan around, and the FAR bonus as an opportunity worth using.

The architect reviews this output, confirms the heritage setback is being read correctly, adds a note about a known planning preference for active street frontage, and approves.

What Buildable Envelope generates: A 3D volume that respects the base setbacks, the heritage facade setback, and the height limit. The corridor bonus increases the volume. Any conflict between the overlay and base zoning gets flagged as a variance for the architect to resolve.

How architects use the output: The envelope becomes the brief for massing. Rather than starting from a box and carving back, you start from what's actually achievable under the layered zoning, and design up from there. Several teams use the envelope specifically to frame conversations with planning authorities early, before the design is developed. It gives those conversations a concrete starting point.

On urban sites with stacked overlays, Buildable Envelope surfaces the layered zoning as a single 3D volume. It's the cleanest way to see what the rules actually add up to.

What changes when the site is a larger commercial parcel?

Larger sites flip the problem. You have space, and the zoning is usually simpler, but the brief is bigger and the consequences of misreading the zoning are larger.

The site: A 6,500 sqm parcel for a commercial office building. Single zoning category. FAR 2.0. Height limit 35 metres. Standard perimeter setbacks. No overlays.

What Analyze Site surfaces: The setbacks, FAR, and height limit, in a clean summary. No special conditions to flag. The opportunities note focuses on the relationship between FAR and footprint: at this FAR, the building can either spread across more of the site at lower height, or concentrate footprint and reach higher. The constraint note flags the perimeter setback as the binding boundary on footprint.

What Buildable Envelope generates: A 3D volume reflecting the maximum buildable area on the site. With straightforward zoning, the envelope is geometrically simple, a setback box at the height limit, with the maximum FAR distributed within.

How architects use the output: On larger commercial sites, the envelope is most useful as a capacity test against the brief. If the client wants 12,000 sqm GFA and the maximum buildable is 13,000 sqm, the design has very little slack. That's a conversation worth having on day one, not three weeks in. Running Buildable Envelope on the imported site delivers the answer in minutes.

On simpler zoning conditions, the agents' value is speed. The envelope arrives fast enough to surface brief-versus-capacity issues before design begins.

How do these agents behave when a site has zoning variances?

Variances are where Buildable Envelope earns its place explicitly. The agent surfaces zoning exceptions and variances rather than absorbing them into the volume. In practice, this is where it differs from a manual envelope calculation.

The site: An infill site in an established residential area. The base zoning sets a 9-metre height limit. The site has a previously approved variance allowing 12 metres for the rear half of the site, granted to a previous owner who didn't build.

What Analyze Site surfaces: Standard setbacks and the base 9-metre height limit. The variance is captured in the zoning data and surfaced as a noted condition alongside the constraints summary. The architect reviews this and confirms the variance is still applicable to the current owner, since variance transferability isn't always automatic.

What Buildable Envelope generates: Two height conditions visible on the envelope. The front portion reflects the 9-metre base limit. The rear portion reflects the 12-metre variance. The variance is flagged in the output, not absorbed silently into the volume.

How architects use the output: The flagged variance is a prompt to verify before design begins. Did the variance carry over with ownership transfer? Is it still valid given any time-based conditions? These are questions the architect needs to answer before treating the higher envelope as available. The agent making the variance visible, rather than baking it in, is what makes this useful. Without the flag, a junior team member might design to the higher envelope without realising the variance needed reverification.

When sites have variances, Buildable Envelope flags them rather than absorbing them. Treat the flag as a prompt to verify, not as automatic permission.

What context do you still need to bring on harder sites?

Adaptive reuse, sites with significant character, and projects where the existing structure is the main story all share one thing: the zoning data alone doesn't tell the whole story.

The site: A 1,400 sqm parcel with an existing two-storey warehouse the client wants to retain and add to. Base zoning permits up to four storeys with standard setbacks. The existing structure is not heritage-listed but is locally valued.

What Analyze Site surfaces: Setbacks, FAR, and height limit for new construction. The agent reads the zoning rules. It doesn't read the existing structure beyond the parcel context, and it doesn't read the local sentiment that makes the structure valuable.

What Buildable Envelope generates: A 3D volume reflecting the four-storey height limit and the setbacks. The envelope assumes a clear site.

What you bring as the architect: The strategy for working with the existing structure. Whether the addition rises above, sits beside, or wraps around. The trade-off between maximising the buildable envelope and preserving the existing massing the neighbourhood values. The structural realities of the existing building. None of that is in the zoning data, and none of it is in the agent's output.

This isn't a limitation of the agents, it's a property of the project type. The agents handle the zoning side cleanly. The contextual reading is yours to bring, and on harder sites, that reading is most of the value you add.

On adaptive reuse and context-heavy projects, the agents handle the zoning baseline. The contextual judgment, the heritage reading, the structural questions, all stay with you.

The site is always different. The process doesn't have to be.

Every site has its own zoning conditions. Tight urban lots, simple commercial parcels, sites with variances, projects where the existing context is the main story. The conditions change every time. What these agents do is standardise the part of the process that doesn't need to be bespoke: reading the zoning data and building it into a 3D volume.

That work belongs to the agents. The design work belongs to you.

Start a new project in Snaptrude and run Analyze Site on your next brief. You'll know within the first session whether it earns a place in your workflow. For the wider picture of how site, massing, and program connect end to end, the 2.5-hour RFP workflow shows the chain in action.

We think it'll earn a place. But the site will tell you more than we can.

‍

Frequently Asked Questions