The 22 skills · Platform skill · Everything only

AGENT SKILL

interactive-maps: what it makes your agent do, and what it refuses.

interactive-maps is the most requested surface in reporting and the easiest place to break every promise the page makes. Tiles phone home, two projections disagree by a few pixels, and the locations that failed to geocode simply do not draw. This skill is the build order that keeps the map as accountable as the table beside it.

The job interactive-maps does

Rule zero is that whatever draws on the map ties to the table and the totals exactly as any other surface does. The build stays self contained: vector shapes only, with the library and the geometry inlined at build time rather than fetched from a content delivery network, which also keeps a strict content policy intact. Reference geometry belongs to the data layer and the fetcher belongs to the engine. One coordinate space is established by construction rather than calibrated afterwards, paint order is treated as an invariant you own, and geocoding is understood as name matching against names that fight back. Anything that did not place goes on a visible unplaced list in every build.

What interactive-maps refuses

A skill is worth what it stops as much as what it produces. These are the refusals written into interactive-maps, taken from the rules and limits in its own file.

• ✓A tile server. A basemap that calls out defeats the self contained page and the content policy, and business geography never needed it.
• ✓A library loaded from a content delivery network. The code and the shapes are inlined at build time.
• ✓A dropped location. Anything that failed to geocode is listed in the build rather than quietly missing from the picture.
• ✓Calibrating a projection by eye. One coordinate space is set by construction, because a few pixels of drift is a permanent, invisible error.

8 lines from the interactive-maps skill file, verbatim

From skills/interactive-maps/SKILL.md, unedited. Kymira skills are plain text you can read in full before you install them, which is the point of shipping a method rather than a binary.

Whatever draws on the map must tie to the table and the totals exactly. A
location that fails to place is **data, never an exception, never silent**:
the geocoder returns a miss, the builder collects misses into an UNPLACED
list and deliberate non-display rows (warehouses, corporate addresses) into
an EXCLUDED list, and **the page prints both**. Every row is in the tables
and totals whether or not it drew a dot. Gate the build on placement counts.
A map that shows 96% of revenue with no note is a wrong number with a nice
projection.

The failure modes interactive-maps addresses

No entry in the catalogue of 51 observed failure modes names interactive-maps as the thing that addresses it. The catalogue records failures we have watched happen, so a skill appears there once a failure it covers has actually been observed and written up. Read the catalogue.

The tier interactive-maps ships in

Both paid tiers are a one-time purchase with lifetime updates, and all sales are final: the files are plain text you keep the moment they arrive, so judge the quality on the free skill and the published doctrine first. The pricing section states exactly what each tier contains.

Installing interactive-maps

A skill is a folder with a SKILL.md file in it. Installing one is copying that folder where your agent looks for skills, with no build step and no dependencies.

mkdir -p ~/.claude/skills
cp -R Kymira/skills/interactive-maps ~/.claude/skills/

Or open the unzipped pack in Claude Code and say install Kymira, which installs every skill in your tier, or double-click the installer for your system. Cursor reads the same skill as a rule file: copy interactive-maps.mdc into .cursor/rules/. Skills are read at startup, so start a new session and ask your agent which skills it has available.

Not for

Last updated 2026-08-13 · Dylan, founder · one of 22 packaged skills, each one a method written down, not a product tour.