# Agentskills Rich

> Convert tabular CSV data into validated JSON records with schema inference. Use when transforming CSV exports into structured JSON, inferring column types, or validating rows against a target schema.

- Skill: `ckodex-labs/agentskills-rich` (Agent Skill, multi-file: 6 files)
- Install (CLI): `npx skillmds@latest add ckodex-labs/agentskills-rich`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ckodex-labs/agentskills-rich/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Data & Analytics
- License: Apache-2.0
- Author: ckodex-labs (https://skillmd.com/u/ckodex-labs)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/ckodex-labs/agentskills-rich

---


# CSV to JSON Conversion

This skill converts CSV files into JSON records with inferred or declared
schemas, and validates each row before emitting output.

## Workflow

1. Read the CSV header and sample the first 100 rows.
2. Infer column types (string, integer, float, boolean, ISO date).
3. Emit one JSON object per row; report rows that fail validation.

## Example

```bash
python scripts/convert.py input.csv --schema references/schema-rules.md
```

Expected output shape:

```json
{ "row": 1, "record": { "id": 42, "name": "example" }, "valid": true }
```

## Verification

Run the fixtures under evals/ and verify each produced record matches the
expected JSON in evals/expected.json. See references/schema-rules.md for the
full inference table (sha256 pinned per release).

## Deep reference

Type-inference edge cases and locale handling live in
references/schema-rules.md — read it before changing inference behavior.