agentleFS
PradnyeshpPradnyeshp/Claude-Skills/.claude/skills/seed/SKILL.md

seed

Skill2 starsChanged 4 months agoMITView on GitHub

Generate realistic seed, fixture, or mock data that matches the project's data models, schema, or types — for local development, demos, or tests. Use when the user says "generate seed data", "create fixtures", "make mock data", "populate the database", "add test data", or "seed the dev db".

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 4

    Distinctive

    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
  1. Very specific:

    Match the project's seeding mechanism: Look for how this project already seeds (a seeds//fixtures/ dir, a prisma/seed.ts, a factory library …

    line 22
  2. Distinctive:

    Make data realistic and coherent: Use plausible values per field (names, emails, dates, enums from their allowed set), respect formats and …

    line 23
  3. Distinctive:

    Never use real personal data or real secrets.

    line 30
  4. Distinctive:

    Don't run destructive operations (truncating tables, resetting the DB) unless the user explicitly asks.

    line 32
  5. Distinctive:

    After writing, report what was generated (models and counts), the command to load it, and any constraints you had to work around.

    line 33

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Schema & model signals

!ls prisma/schema.prisma schema.sql migrations models 2>/dev/null

!grep -rilE "schema|model|entity|@Entity|class .*\\(Base\\)|createTable|Schema\\(" --include='*.ts' --include='*.js' --include='*.py' --include='*.prisma' --include='*.rb' . 2>/dev/null | grep -vE 'node_modules|\.git|dist|build|vendor' | head -20

Instructions

Generate seed/fixture data for this project per $ARGUMENTS (what to seed and how many). If the target model/table isn't clear, find the schema first and confirm scope before generating.

Method

  1. Read the real schema. Locate the models/tables/types (Prisma schema, SQL DDL, ORM models, TypeScript types, pydantic/dataclasses) and extract every field: type, nullability, defaults, enums, unique constraints, and foreign keys. Generate against the actual shape — don't guess columns.
  2. Match the project's seeding mechanism. Look for how this project already seeds (a seeds//fixtures/ dir, a prisma/seed.ts, a factory library like factory_boy/FactoryBot/Faker, a db:seed script). Extend that mechanism and reuse its style rather than introducing a new one.
  3. Make data realistic and coherent. Use plausible values per field (names, emails, dates, enums from their allowed set), respect formats and length limits, and keep relationships valid — generate parents before children and use real foreign keys so referential integrity holds.
  4. Honor constraints. Unique fields get unique values; required fields are always set; enums only use defined variants; numeric/date ranges stay sensible. Make output deterministic (seed the RNG / use a fixed seed) so runs are reproducible.
  5. Write to the conventional location (the existing seed script or fixtures dir) and wire it to the project's seed command if one exists. Default to a modest volume unless the user asked for a specific count.

Rules

  1. Generate strictly against the real schema — every field, type, and constraint comes from the code, not assumption. Flag anything you couldn't determine rather than inventing it.
  2. Never use real personal data or real secrets; synthesize fake-but-plausible values, and clearly fake credentials for any auth fields.
  3. Preserve referential integrity — no dangling foreign keys, no constraint violations the database would reject.
  4. Don't run destructive operations (truncating tables, resetting the DB) unless the user explicitly asks; default to additive, idempotent seeds where possible.
  5. After writing, report what was generated (models and counts), the command to load it, and any constraints you had to work around.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.