cardstack/boxel-skills/rich-markdown-reports
- community
Checking that the SKILL.md file is still in the repository.
Use when a request calls for a report, summary, briefing, or dashboard-style document — especially when you'd otherwise define a new card type to present it. Compose it as a rich Boxel markdown (.md) file that embeds the relevant cards (custom or off-the-shelf) instead of authoring a bespoke CardDef or writing plain prose.
Install
npx vskill@latest install cardstack/boxel-skills/rich-markdown-reportsSKILL.md
Rich Markdown Reports
_For a report, reach for Rich Markdown that embeds existing cards — not a new card definition built just to present it._
When a user asks for a report, summary, or dashboard-style document, the tempting move is to design a new card type (a CardDef) to hold it. For most reports, don't. Compose the report as a Rich Markdown `.md` file that embeds the cards it's about — custom types the user already has, or off-the-shelf catalog cards. You get a live, rich document without inventing a one-off schema, and the embeds stay current as the underlying cards change.
The rule
When a task calls for a report, summary, meeting notes, briefing, or any write-up of one or more cards:
- Default to Rich Markdown, not a new card definition. Reach for a
.mdfile that embeds existing cards before you consider authoring a bespokeCardDefto represent the report. - Embed the cards the report is about with BFM directives instead of describing them in prose — custom or off-the-shelf cards both work.
- Reference live data rather than copying values into the text, so the report stays current as the cards change.
- Structure it with headings so it's navigable.
Write it as a `.md` file in a realm — do not paste it into the chat. The card and file embeds only render as live cards when the markdown lives as a file in a realm; the same BFM typed into a chat reply stays raw text and the embeds don't resolve. Save it to a realm the user can write to — see Pair with below for how to create the file. A report that never becomes a realm file hasn't used this skill.
Pair with
- `boxel-flavored-markdown` — the directive grammar your embeds use (
:card/::card/::file, embed formats, fenced renderers). Pick the format that fits the report. - `source-code-editing` — the
run-realm-codetool used to write the file to the realm. It works the same for a.mdfile as for code, so use it here even though a report isn't source code.
Don't use for
- A reusable, structured, or interactive artifact. If the deliverable needs a typed schema, edit forms, or its own behavior — not just a document that presents other cards — a
CardDefis the right tool, not Rich Markdown. - Content that references no Boxel cards or files — plain markdown is fine there.
- BFM grammar edge cases or a card's rendering internals — that's the
boxel-flavored-markdownskill.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/rich-markdown-reports/SKILL.md
Popularity
- Repository stars
- 1
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/cardstack/boxel-skills/rich-markdown-reports)