Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit and migrate an existing Markdown renderer to Markstream while preserving custom renderers, security policy, streaming behavior, and explicit parity gaps.
.claude/skills/sickn33-markstream-migration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 83% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 25% | 0% |
Replace an existing Markdown renderer without silently dropping transforms, custom components, URL policy, raw-HTML behavior, or streaming semantics. Read references/adoption-checklist.md first.
Use when replacing react-markdown, markdown-it, marked, or another renderer; migrating node renderers; or choosing between Markstream content, smooth streaming, and nodes.
Before changing dependencies or source files, inspect the existing package manager and project conventions, preview the intended edits, and obtain explicit user approval.
content with smooth streaming for ordinary token streams. Use nodes only for worker parsing, shared AST ownership, or structural transforms.tsx// Before: // import ReactMarkdown from 'react-markdown' // return <ReactMarkdown>{markdown}</ReactMarkdown> import MarkdownRender from 'markstream-react' import 'markstream-react/index.css' export function AssistantAnswer({ markdown, isDone, }: { markdown: string isDone: boolean }) { return ( <MarkdownRender content={markdown} final={isDone} fade={isDone} typewriter={!isDone} smoothStreaming={isDone ? false : 'auto'} htmlPolicy="safe" /> ) }
Do not weaken sanitization for screenshot parity. Review dependencies, raw HTML, URL transforms, and trust boundaries explicitly.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 12,852 | 12,636 | -2% | 1 | 1 | 0% | 2,434 | 3,036 | +25% | 0 | 0 | — |
case-06 | pass→pass | 16,257 | 11,818 | -27% | 1 | 1 | 0% | 2,775 | 2,733 | -2% | 0 | 0 | — |
case-01 | fail→fail | 13,577 | 9,049 | -33% | 1 | 1 | 0% | 2,700 | 2,290 | -15% | 0 | 0 | — |
case-02 | fail→fail | 18,784 | 10,278 | -45% | 1 | 1 | 0% | 3,527 | 2,451 | -31% | 0 | 0 | — |
case-03 | fail→fail | 16,117 | 10,037 | -38% | 1 | 1 | 0% | 3,147 | 2,275 | -28% | 0 | 0 | — |
case-04 | pass→pass | 18,220 | 11,176 | -39% | 1 | 1 | 0% | 3,585 | 2,979 | -17% | 0 | 0 | — |
case-07 | fail→pass | 14,754 | 7,126 | -52% | 1 | 1 | 0% | 2,234 | 1,744 | -22% | 0 | 0 | — |
case-08 | pass→pass | 11,717 | 8,453 | -28% | 1 | 1 | 0% | 2,028 | 2,062 | +2% | 0 | 0 | — |
case-09 | fail→fail | 13,401 | 8,220 | -39% | 1 | 1 | 0% | 2,234 | 1,740 | -22% | 0 | 0 | — |
case-10 | fail→pass | 5,535 | 6,334 | +14% | 1 | 1 | 0% | 916 | 1,677 | +83% | 0 | 0 | — |
case-11 | pass→pass | 13,464 | 7,218 | -46% | 1 | 1 | 0% | 2,465 | 1,857 | -25% | 0 | 0 | — |
case-12 | pass→pass | 14,433 | 7,167 | -50% | 1 | 1 | 0% | 2,444 | 1,659 | -32% | 0 | 0 | — |
case-13 | pass→pass | 11,352 | 6,483 | -43% | 1 | 1 | 0% | 1,866 | 1,547 | -17% | 0 | 0 | — |
case-14 | pass→pass | 14,196 | 7,294 | -49% | 1 | 1 | 0% | 2,312 | 1,740 | -25% | 0 | 0 | — |
case-15 | fail→pass | 7,646 | 4,974 | -35% | 1 | 1 | 0% | 1,525 | 1,528 | +0% | 0 | 0 | — |
case-16 | pass→pass | 9,689 | 3,839 | -60% | 1 | 1 | 0% | 1,682 | 1,217 | -28% | 0 | 0 | — |
case-17 | pass→pass | 15,110 | 7,750 | -49% | 1 | 1 | 0% | 2,342 | 1,892 | -19% | 0 | 0 | — |
case-18 | pass→pass | 12,302 | 6,096 | -50% | 1 | 1 | 0% | 1,994 | 1,638 | -18% | 0 | 0 | — |
case-19 | pass→pass | 14,047 | 10,277 | -27% | 1 | 1 | 0% | 2,480 | 2,299 | -7% | 0 | 0 | — |
case-20 | pass→pass | 13,141 | 8,014 | -39% | 1 | 1 | 0% | 2,232 | 1,887 | -15% | 0 | 0 | — |
case-21 | fail→fail | 10,872 | 8,566 | -21% | 1 | 1 | 0% | 1,873 | 1,945 | +4% | 0 | 0 | — |
case-22 | fail→pass | 7,669 | 6,310 | -18% | 1 | 1 | 0% | 1,443 | 1,785 | +24% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.