report-artifact
Skill0 starsChanged 4 months agoMITView on GitHub
The shared visual design system for the kit's beautiful, self-contained HTML report Artifacts — screen shortlist, portfolio evaluation, watchlist status, research brief, premarket/EOD digest. Use whenever another skill renders a report as a claude.ai Artifact so every report shares one calm, disciplined, scannable identity with provenance labels and the SEBI disclaimer. Read-only presentation only.
In this repository8 other files it gives its agents.
- autopilotSkill
- checkSkill
- india-dataSkill
- opportunity-screenSkill
- portfolio-evalSkill
- portfolio-watchSkill
- researchSkill
- watchlist-monitorSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Self-contained, CSP-safe: No external fonts, scripts, images, or fetch/XHR — the artifact CSP blocks them.
- Distinctive: line 17
Compliance copy: No price targets, no "assured/guaranteed/risk-free/target-return", no performance claims, no SEBI reg number / exchange …
- Distinctive: line 49
Never compute or invent a number in the artifact.
- Distinctive: line 50
Never omit the disclaimer footer or the delayed-data labels.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Report Artifact — shared design system (survival-first)
Render reports as a self-contained claude.ai Artifact (single HTML file) only on surfaces that support Artifacts; otherwise emit a clean text digest. The report presents numbers the deterministic gate/helpers already computed — it never computes or asserts a figure itself, and never adds a number the data didn't carry.
Hard rules (non-negotiable)
- Self-contained, CSP-safe. No external fonts, scripts, images, or fetch/XHR — the artifact CSP blocks them. Inline all CSS; use a system font stack (below); draw any chart with inline
<canvas>+ JS, not remote libraries. - Theme-aware. Define tokens on
:root, redefine under@media (prefers-color-scheme: dark)and under:root[data-theme="dark"]/:root[data-theme="light"](the viewer's toggle stampsdata-themeand must win both directions). Style through tokens, never hard-coded colors. - Figures line up.
font-variant-numeric: tabular-numson every column of money/percent/shares. Wide tables live in anoverflow-x: autocontainer so the body never scrolls sideways. - Provenance on every number. Show
delay_class(live / delayed / manual) andas_ofnear each figure or as a per-section stamp — delayed data is always labeled delayed. - Verdict = semantic state, not accent. Encode state in form: a pill/chip with its own color, separate from the accent hue — HOLD/TRADE = good, TRIM/caution = warn, EXIT/NO-TRADE/STAND-DOWN = critical/neutral. Scannable at a glance.
- Mandatory disclaimer footer — the verbatim SEBI-safe disclaimer (from
scripts/disclaimer.py) at the foot of every report. - Compliance copy. No price targets, no "assured/guaranteed/risk-free/target-return", no performance claims, no SEBI reg number / exchange logos. Default framing is NO-TRADE.
Design plan (tokens)
Calm, disciplined, ledger-like — the opposite of a casino/finfluencer look. Neutrals carry a slight cool ink bias; one restrained accent (deep teal); semantic greens/ambers/reds reserved for verdict state.
:root{
--paper:#f7f8f7; --ink:#12201f; --muted:#5c6b69; --line:#e2e6e4; --card:#ffffff;
--accent:#0f6f6b; /* deep teal — the one bold hue, used sparingly */
--good:#2f7a4f; --warn:#a86b12; --bad:#b23b3b; --neutral:#6b7472; /* semantic verdicts */
--mono:ui-monospace,"SF Mono",Menlo,Consolas,monospace;
--sans:system-ui,-apple-system,"Segoe UI",Roboto,sans-serif;
--serif:ui-serif,Georgia,"Times New Roman",serif; /* display headings, used with restraint */
}
@media (prefers-color-scheme:dark){:root{
--paper:#0e1413; --ink:#e7edea; --muted:#93a19e; --line:#233029; --card:#141c1a;
--accent:#3fb3ad; --good:#57b07e; --warn:#d29b4a; --bad:#e0736f; --neutral:#8a938f;
}}
:root[data-theme="dark"]{--paper:#0e1413;--ink:#e7edea;--muted:#93a19e;--line:#233029;--card:#141c1a;--accent:#3fb3ad;--good:#57b07e;--warn:#d29b4a;--bad:#e0736f;--neutral:#8a938f;}
:root[data-theme="light"]{--paper:#f7f8f7;--ink:#12201f;--muted:#5c6b69;--line:#e2e6e4;--card:#ffffff;--accent:#0f6f6b;--good:#2f7a4f;--warn:#a86b12;--bad:#b23b3b;--neutral:#6b7472;}Body text in --sans near 65ch; section headings in --serif with text-wrap:balance; all data in --mono tabular. Uppercase eyebrow labels get slight letter-spacing. Respect prefers-reduced-motion; give focus a visible ring.
Layout — summary first, then detail
Open with a status banner (the one thing that needs attention: regime posture, or the dominant verdict), then the detail. Per report:
- Opportunity screen — a shortlist table:
symbol · entry · stop · shares · target(R) · net break-even · dominant reason, each row a verdict pill. If empty: a single calm "No trade — dominant gate: …" card. - Portfolio eval — total value + concentration/health callouts up top; a composition table with horizontal weight bars (canvas or CSS), unrealized P&L% colored by sign, and a HOLD/TRIM/EXIT pill per name.
- Watchlist — setups grouped by status (triggered → armed → extended → invalidated/aged-out), each with pivot · price · age; extended/invalidated visually de-emphasized.
- Research brief — sections (snapshot · risks/red-flags first · quality · valuation · technical) with a red-flags table and a sources list.
- Premarket / EOD digest — regime posture banner + holdings-needing-action + up to 3 opportunities, compact.
Never
- Never compute or invent a number in the artifact; render only what the gate/helpers produced, with its provenance.
- Never omit the disclaimer footer or the delayed-data labels. Never use forbidden performance/target language.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

