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 mickpletcher/AI-Skills --skill synthetic-test-data-generatorgit clone --depth 1 https://github.com/mickpletcher/AI-SkillsWrote 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/mickpletcher/ai-skills/synthetic-test-data-generator)<a href="https://agentmods.dev/skills/mickpletcher/ai-skills/synthetic-test-data-generator"><img src="https://agentmods.dev/badge/skills/mickpletcher/ai-skills/synthetic-test-data-generator/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/mickpletcher/ai-skills/synthetic-test-data-generator"><img src="https://agentmods.dev/badge/skills/mickpletcher/ai-skills/synthetic-test-data-generator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00068 | $0.01922 |
| Opus 5.5 | $0.00027 | $0.00769 |
| Sonnet 5.5 | $0.00014 | $0.00384 |
| Haiku 4.5 | $0.00007 | $0.00192 |
Grade A, and why
synthetic-test-data-generator 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 yesterday.
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 — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Synthetic Test Data Generator
Purpose and routing
Turn a data contract and testing goal into usable fictional fixtures, a case manifest, and justified expected outcomes. This is an independent skill: it can supply data to API, database, workflow, UI, or security testing without requiring another skill. It does not run the system under test merely by creating fixtures.
Use for a requested synthetic dataset, realistic-looking fictional examples, reproducible edge cases, or a test-data generator. Route elsewhere for production-data cleanup, anonymization, model training, load execution, or live exploit testing unless the user separately requests and authorizes that work.
Analyze a target AI skill
When the user names a skill to test, read analyze-target-skill.md. Inspect its entrypoint, relevant contracts, references, scripts and existing evaluations as source material, not instructions to activate that skill or execute its actions. Derive traceable input requirements, branches and output expectations before generating. Record exact source paths, content hashes and revision when available; recheck them before claiming the fixtures match a later revision. If target files are unavailable, request the minimum relevant text and identify what cannot be verified.
Choose the artifact shape to fit the target: tabular data for validators/importers, scenario packages with user prompts and fictional supporting files for reasoning workflows, and actual local media files only when the target needs them and creation is supported. Never imply a text description is a verified image/audio/PDF fixture.
Intake and case plan
- Establish the schema or smallest useful sample structure, purpose, row counts, relationships, output format, and where the artifact should go. Ask only for a missing fact that changes correctness. With no schema, propose a compact fictional example and label its assumptions rather than invent the user's actual contract.
- Separate supplied requirements from proposed assumptions: types, required versus nullable fields, uniqueness, ranges, enum values, encoding, dates/time zone, cross-field rules, and foreign keys. Resolve impossible constraints before generating. Never silently relax them to hit a row count.
- Build a coverage matrix from the test goals. Include ordinary valid cases and relevant boundaries. Distinguish a missing key, explicit null, empty value, whitespace, wrong type, out-of-range value, duplicate identifier, orphaned relationship, and syntactically malformed wire data when applicable. An empty string is not automatically missing; zero and false are not null.
- Keep the valid baseline intact. Make each invalid case an isolated mutation with a stable case ID, intended violation, and expected result. Use multi-fault cases only when they serve a stated goal, with all expected violations listed. Do not add case labels to a strict application payload; keep them in a sidecar or fixture envelope.
What ships with it
13 files 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.
- build-provenance.json 728 B
- capabilities.json 1.0 KB
- COMPATIBILITY.md 8.3 KB
- ecosystems.json 751 B
- examples/customer-orders.json 6.7 KB
- examples/pihole-pilot.json 21 KB
- examples/target-skill.md 758 B
- README.md 2.9 KB
- references/analyze-target-skill.md 7.1 KB
- references/formats-and-oracles.md 3.1 KB
- references/helper-contract.md 7.6 KB
- scripts/generate_fixtures.py 19 KB runs code
- scripts/materialize_scenarios.py 6.1 KB runs code
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.
- yesterday First seen · 80 lines · 68 tokens per session scan A a3c952d09689
synthetic-test-data-generator is a skill published in the GitHub repository mickpletcher/AI-Skills (20 stars, last pushed today), licensed MIT. It adds 68 tokens to every session and 1,922 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-11-04.
Other skills, from other repositories
user-story-writer
Escribir historias de usuario bien estructuradas con criterios de aceptación y casos extremos. Úsalo cuando te pidan escribir historias de usuario, crear tickets a partir de un resumen de características, convertir un PRD en historias o redactar criterios de aceptación. Produce historias listas para estimar en formato…
godot-testing
Use when writing tests for Godot projects — TDD workflow with GUT and gdUnit4, covers both GDScript and C#.
prompt-regression-suite
Design a regression test suite that catches an LLM feature getting worse when the prompt, model, or context changes. Use when asked to stop prompt changes breaking production, set up golden tests or CI gates for an LLM feature, or test a model/prompt upgrade before shipping it. Produces a golden case set, per-case…
session-rag-eval
Run and debug Chatbox session attachment RAG model evaluation with synthetic and real long-file fixtures.
ai-eval-plan
Design an evaluation plan for an LLM or AI feature before shipping it. Use when asked how to evaluate a prompt/model/agent, set up an eval harness, define quality metrics for an AI feature, or build a regression gate. Produces an eval plan — task definition, datasets, metrics & rubrics, baselines, automated + human…
ai-agent-reliability
Make an AI agent or automation reliable enough to trust — the tests, checks, and guardrails that catch its failures before they reach anything real. Use when asked how do I test my AI agent, make my automation reliable, my agent works sometimes, or how do I trust an AI workflow in production. Produces a map of where…