The 22 skills · Platform skill · Everything only

AGENT SKILL

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

interactive-reports adds the first clicks every viewer expects, without giving up a tie. A filtered view that does not re-tie is a wrong number wearing an interactive costume, so every figure derives in script from the one data block that built the page, and the checks recompute with the filter.

The job interactive-reports does

The interactions are table stakes: pick last thirty days, pick a region, click a bar and expect the rest to follow. This skill ships them inside a single static file. One baked data block is the only source of truth, and every total, chart and check derives from it, so no pre-rendered number can be left behind by a click. Date ranges get explicit boundary handling, filters and cross filtering state what they are showing, drill down and drill through keep their lineage, and a show me the rows affordance closes the trust loop between a figure and the records under it. The manifest that ships with the page has to satisfy an auditor, and the mobile view is treated as the executive's actual reading surface.

What interactive-reports refuses

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

• ✓A second copy of the data. A value typed into the HTML and also derivable from the block gives the page two sources of truth, and one of them will lie after the first click.
• ✓Checks that ignore the filter. When a filter changes what is on screen, the checks recompute over that slice or say plainly that the ties shown are for the full period.
• ✓Interactivity as load bearing. The initial view is a complete report before any script runs, because a stalled script must still leave a correct page.
• ✓Hiding rows with script. Per viewer restriction is an access control question, and doing it in the browser is theatre.

8 lines from the interactive-reports skill file, verbatim

From skills/interactive-reports/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.

All interactivity is client-side over **the same baked JSON block that
built the page**. Every figure, chart, total, and check **derives in script
from that one block** — never a second copy, never a pre-rendered number
that a filter forgets to update. If a value is typed into the HTML and also
derivable from data, the page has two sources of truth and one of them will
lie after the first click.

**And the checks are part of the view.** When a filter changes what's on

The failure modes interactive-reports addresses

These entries in the catalogue of 51 observed failure modes name this skill as what addresses them. Every one was seen in a real build or in our own audits, and none was invented.

The tier interactive-reports 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-reports

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-reports ~/.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-reports.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 2026-08-29 · Dylan, founder · one of 22 packaged skills, each one a method written down, not a product tour.