Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add OutlineDriven/outline-driven-development --skill dataflow-diagramgit clone --depth 1 https://github.com/OutlineDriven/outline-driven-developmentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/outlinedriven/outline-driven-development/dataflow-diagram)<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/dataflow-diagram"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/dataflow-diagram/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/dataflow-diagram"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/dataflow-diagram.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.01171 |
| Opus 5 | $0.00017 | $0.00585 |
| Sonnet 5 | $0.00007 | $0.00234 |
| Haiku 4.5 | $0.00003 | $0.00117 |
Grade A, and why
dataflow-diagram scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dataflow diagram
Contract
| Field | Bound contract |
|---|---|
| Trigger | The user asks to visualize data movement, ETL or ELT, lineage, transformation stages, custody, governance, stores, sources, or consumers. |
| Authority | Reversible local: writes only the named local artifacts below; rollback is deleting them. No remote mutation. |
| Side effect | Writes a typed dataflow JSON specification and one self-contained interactive HTML artifact, with optional bounded visual-evidence sidecars. |
| Done | The diagram represents the requested movement and custody facts without inferred causality, validates against both schemas, passes all render and atomic-delivery checks, and has a hash-bound receipt plus truthful visual-review status. |
Inputs
The user must supply the data contracts and movement facts to depict: sources, stores, consumers, transformation stages, custody owners, and the flows between them. Optional inputs are governance labels, lineage annotations, and a request for visual-evidence sidecars. Do not infer a flow, stage, custody assignment, or causal relationship the user did not state; ask for the missing fact rather than inventing it.
Procedure
- Bound scope before mutation: confirm the requested movement and custody facts, and list the exact local files to write (the dataflow JSON specification and one self-contained interactive HTML artifact, plus any requested visual-evidence sidecars). Write nothing else. Done when: the file set is listed and confirmed before any mutation.
- Model the dataflow as stages, nodes, flows, and semantic kinds using only the user-supplied facts. Record each node's kind (source, store, transformation, consumer, governance checkpoint) and each flow's source, target, and custody owner as stated. Add no flow, stage, or causal edge the user did not provide. Done when: the model contains only user-supplied facts with no inferred edges.
- Serialize the model into the typed dataflow JSON specification. Validate it strictly against the dataflow schema and the common schema: every required field present, every semantic kind from the allowed set, every flow resolvable to declared nodes, and no extra or unknown fields accepted. Done when: the JSON specification validates strictly against both schemas.
- Render one self-contained interactive HTML artifact from the validated specification. The artifact must load with no external network or runtime dependency, render the stages and flows with their semantic kinds and custody owners, and let a reader traverse the movement interactively. Done when: the HTML artifact loads with no external dependency and renders all stages and flows interactively.
- Run the render check (the artifact renders the full requested movement and custody facts) and the atomic-delivery check (the JSON specification and the HTML artifact are each complete and usable on their own). If either fails, fix the artifact or specification and re-run; do not declare done on a partial pass. Done when: both the render and atomic-delivery checks pass.
- Capture a hash-bound receipt: record the SHA-256 of the delivered JSON specification and HTML artifact, and the schema versions validated against. Record a truthful visual-review status stating whether a human visually confirmed the rendered artifact, was not yet reviewed, or reviewed and found defects; never mark unreviewed output as confirmed. Done when: the receipt records both hashes, schema versions, and a truthful visual-review status.
- If visual-evidence sidecars were requested, produce only the bounded set asked for (for example, a screenshot or rendered snapshot of the artifact) and name them in the receipt. Do not generate unrequested sidecars. Done when: requested sidecars are produced and named in the receipt, or no sidecars are produced when none were requested.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago Changed · -26 tokens per session 540c167d5305
- 5d ago First seen · 41 lines · 60 tokens per session scan A 96d0f9c9c391
dataflow-diagram is a skill published in the GitHub repository OutlineDriven/outline-driven-development (52 stars, last pushed 3d ago), licensed Apache-2.0. It adds 34 tokens to every session and 1,171 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
outlines
Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library.
outlines
Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library.
outlines
Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library.
outlines
Outlines: structured JSON/regex/Pydantic LLM generation.
outlines
Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library.
analysis-artifacts
Use when the user requests a deep dive, exploratory analysis, or data analysis on BigQuery. Not for credential, publish, deploy, or irreversible changes.