Agent skill

PocketBase Collections

Collection and schema design for PocketBase. Use when creating collections, designing schemas, adding fields, setting up relations, or choosing between base/auth/view collection types. Prevents wrong field types, documents zero-default behavior, and covers relation cascading.

Stars 23,776
Forks 2,298

Install this agent skill to your Project

npx add-skill https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/pocketbase/pb-collections

SKILL.md

PocketBase Collection & Schema Design

Collection Types

Base Collection

Standard data collection. System fields: id, created, updated.

Auth Collection

Extends base with authentication. Additional system fields: email, emailVisibility, verified, password, tokenKey.

Cannot delete system fields. Can disable email/password auth in collection options.

View Collection

Read-only, backed by a SQL SELECT query. No create/update/delete. Fields are auto-detected from the query. Useful for aggregations, joins, and computed views.

sql
-- Example: view collection query
SELECT p.id, p.title, COUNT(c.id) as comments_count
FROM posts p LEFT JOIN comments c ON c.post = p.id
GROUP BY p.id

View collections support API rules (list/view only) and can be used in relations.

Field Types

Type Go type Zero default Notes
text string "" min/max length, regex pattern
editor string "" Rich text (sanitized HTML)
number float64 0 min/max, noDecimal option
bool bool false
email string "" Auto-validated format
url string "" Auto-validated format
date string "" ISO 8601 (2024-01-01 00:00:00.000Z)
select string/[]string ""/[] values list, maxSelect
file string/[]string ""/[] maxSelect, maxSize, mimeTypes
relation string/[]string ""/[] collectionId, cascadeDelete, maxSelect
json any null Only type that can be null! maxSize
autodate string auto onCreate/onUpdate modifiers
password string "" Stored hashed, never returned in API

Critical: all types default to their zero value, NOT null. Only json fields can be null.

Field Modifiers

Use in collection schema definitions:

  • required — field cannot be empty/zero
  • unique — unique constraint (composite via unique indexes)
  • presentable — included in relation display
  • hidden — excluded from API responses unless explicitly requested
  • :autogenerate — for text fields: auto-generate value (e.g., slug from other field)

Relation Patterns

One-to-many

posts.author -> users (maxSelect: 1)

Each post has one author. Query posts by author: author = "USER_ID".

Many-to-many

posts.tags -> tags (maxSelect: 0, meaning unlimited)

Multi-select relation. Filter: tags ?= "TAG_ID" (contains).

Back-relations

No explicit back-relation field needed. Use @collection.posts.author in API rules or expand from either side:

GET /api/collections/users/records/USER_ID?expand=posts_via_author

Cascade Delete

Set cascadeDelete: true on the relation field. When the referenced record is deleted, all records pointing to it are also deleted. Default is false (sets to empty string).

Self-referencing

A collection can reference itself:

categories.parent -> categories (maxSelect: 1)

Indexes

  • Created in the collection settings (not field-level)
  • Format: CREATE [UNIQUE] INDEX idx_name ON collection (field1, field2)
  • Unique indexes enforce composite uniqueness
  • Partial indexes: CREATE INDEX ... WHERE condition
  • Indexes on relation fields improve join performance

Auth Collection Specifics

OAuth2

Enable per-provider in collection settings. Each provider needs client ID + secret. PocketBase handles the full OAuth2 flow.

OTP (One-Time Password)

Enable in auth collection settings. Sends code via email. Configure otp.enabled, otp.duration, otp.length.

MFA (Multi-Factor Authentication)

Enable in auth collection settings. Requires a second factor after primary auth. mfa.enabled, mfa.duration, mfa.rule (filter to determine which users need MFA).

Password Auth

Enabled by default. Can customize minPasswordLength. Can disable entirely if using only OAuth2/OTP.

Auth Options

  • authToken.duration — token lifetime (seconds)
  • passwordAuth.enabled — toggle email/password
  • passwordAuth.identityFields — fields used for login (default: email; can add username)
  • oauth2.enabled — toggle OAuth2
  • otp.enabled — toggle OTP

Best Practices

  1. Prefer select over bool when there might be more than 2 states in the future
  2. Use relation not text for foreign keys — you get cascade, expand, and type safety
  3. json fields are schemaless — use sparingly, prefer typed fields
  4. Name collections in lowercase snake_case (e.g., blog_posts, user_profiles)
  5. Index early — add indexes for any field used in filters or sorts
  6. maxSelect: 1 on relations returns a string ID; maxSelect: >1 or 0 returns an array

Expand your agent's capabilities with these related and highly-rated skills.

davila7/claude-code-templates

verl-rl-training

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

23,776 2,298
Explore
davila7/claude-code-templates

openrlhf-training

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

23,776 2,298
Explore
davila7/claude-code-templates

gguf-quantization

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.

23,776 2,298
Explore
davila7/claude-code-templates

Claude Code Guide

Master guide for using Claude Code effectively. Includes configuration templates, prompting strategies "Thinking" keywords, debugging techniques, and best practices for interacting with the agent.

23,776 2,298
Explore
davila7/claude-code-templates

qdrant-vector-search

High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.

23,776 2,298
Explore
davila7/claude-code-templates

behavioral-modes

AI operational modes (brainstorm, implement, debug, review, teach, ship, orchestrate). Use to adapt behavior based on task type.

23,776 2,298
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results