Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generates a Mermaid sequence diagram showing how data moves between components. Use when tracing request flows or documenting data transformation pipelines.
.claude/skills/athola-data-flow/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -26% | 0% |
Generate a Mermaid sequence diagram showing how data moves between components in a codebase.
cartograph:call-chain)cartograph:workflow-diagram)Dispatch the codebase explorer agent:
Agent(cartograph:codebase-explorer)
Prompt: Explore [scope] and return a structural model.
Focus on function calls, data transformations, and
inter-module communication for a data flow diagram.Transform the structural model into a Mermaid sequence diagram.
Rules for data flow diagrams:
sequenceDiagram for request/response flows->> for calls,-->> for returns
activate/deactivate for long-running operationsNote over for data transformationsalt/else for conditional flowsExample output:
mermaidsequenceDiagram participant User participant Command as /commit participant Sanctum as sanctum.commit participant Leyline as leyline.git_platform participant Git User->>Command: /commit Command->>Sanctum: generate_message() Sanctum->>Leyline: get_staged_changes() Leyline->>Git: git diff --cached Git-->>Leyline: diff output Leyline-->>Sanctum: structured changes Note over Sanctum: Classify change type Sanctum-->>Command: commit message Command->>Git: git commit -m "..."
Call the Mermaid Chart MCP to render:
mcp__claude_ai_Mermaid_Chart__validate_and_render_mermaid_diagram
prompt: "Data flow diagram of [scope/feature]"
mermaidCode: [generated syntax]
diagramType: "sequenceDiagram"
clientName: "claude-code"If rendering fails, fix syntax and retry (max 2 retries).
Show the rendered diagram with a brief description of the flow depicted (2-3 sentences).
sequenceDiagram syntax generated with participantsand at least one ->> call arrow
mcp__claude_ai_Mermaid_Chart__validate_and_render_mermaid_diagramcalled with diagramType: "sequenceDiagram" and returns without error (or retry attempted on first failure)
they are aggregated and the aggregation noted
alongside the rendered diagram
-->>) are present for any request that hasa corresponding response in the traced flow
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,588 | 9,077 | -22% | 1 | 1 | 0% | 1,922 | 1,155 | -40% | 0 | 0 | — |
case-02 | fail→fail | 13,921 | 5,768 | -59% | 1 | 1 | 0% | 2,285 | 1,019 | -55% | 0 | 0 | — |
case-03 | fail→fail | 14,051 | 5,459 | -61% | 1 | 1 | 0% | 2,411 | 985 | -59% | 0 | 0 | — |
case-04 | fail→pass | 12,219 | 3,544 | -71% | 1 | 1 | 0% | 1,928 | 1,266 | -34% | 0 | 0 | — |
case-05 | fail→pass | 11,487 | 4,335 | -62% | 1 | 1 | 0% | 1,872 | 1,393 | -26% | 0 | 0 | — |
case-06 | fail→pass | 11,910 | 6,348 | -47% | 1 | 1 | 0% | 2,062 | 1,605 | -22% | 0 | 0 | — |
case-07 | pass→pass | 13,464 | 4,903 | -64% | 1 | 1 | 0% | 2,129 | 1,341 | -37% | 0 | 0 | — |
case-08 | pass→pass | 8,031 | 3,734 | -54% | 1 | 1 | 0% | 1,331 | 1,297 | -3% | 0 | 0 | — |
case-09 | fail→fail | 16,026 | 7,440 | -54% | 1 | 1 | 0% | 3,127 | 1,203 | -62% | 0 | 0 | — |
case-10 | fail→pass | 18,453 | 1,770 | -90% | 1 | 1 | 0% | 1,655 | 1,024 | -38% | 0 | 0 | — |
case-11 | fail→pass | 12,057 | 3,429 | -72% | 1 | 1 | 0% | 1,797 | 1,329 | -26% | 0 | 0 | — |
case-20 | pass→fail | 3,711 | 1,433 | -61% | 1 | 1 | 0% | 550 | 891 | +62% | 0 | 0 | — |
case-12 | pass→pass | 10,515 | 4,789 | -54% | 1 | 1 | 0% | 1,883 | 1,452 | -23% | 0 | 0 | — |
case-13 | pass→pass | 6,784 | 1,625 | -76% | 1 | 1 | 0% | 976 | 988 | +1% | 0 | 0 | — |
case-14 | pass→pass | 9,739 | 4,790 | -51% | 1 | 1 | 0% | 1,617 | 1,522 | -6% | 0 | 0 | — |
case-15 | pass→pass | 12,070 | 4,906 | -59% | 1 | 1 | 0% | 1,986 | 1,547 | -22% | 0 | 0 | — |
case-16 | pass→pass | 12,622 | 7,789 | -38% | 1 | 1 | 0% | 2,014 | 2,127 | +6% | 0 | 0 | — |
case-17 | pass→fail | 12,426 | 42,718 | +244% | 1 | 1 | 0% | 2,023 | 3,715 | +84% | 0 | 0 | — |
case-18 | fail→fail | 9,731 | 2,292 | -76% | 1 | 1 | 0% | 1,493 | 990 | -34% | 0 | 0 | — |
case-19 | pass→pass | 12,931 | 7,788 | -40% | 1 | 1 | 0% | 2,027 | 1,911 | -6% | 0 | 0 | — |
case-21 | fail→fail | 10,651 | 2,015 | -81% | 1 | 1 | 0% | 1,678 | 1,107 | -34% | 0 | 0 | — |
case-22 | fail→pass | 17,121 | 6,768 | -60% | 1 | 1 | 0% | 2,668 | 1,727 | -35% | 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, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +18 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
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.