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.
git clone --depth 1 https://github.com/JoaoEquer/OficinaWrote 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/commands/joaoequer/oficina/crud)<a href="https://agentmods.dev/commands/joaoequer/oficina/crud"><img src="https://agentmods.dev/badge/commands/joaoequer/oficina/crud/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/commands/joaoequer/oficina/crud"><img src="https://agentmods.dev/badge/commands/joaoequer/oficina/crud.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 | $0.00019 | $0.00697 |
| Opus 5.5 | $0.00008 | $0.00279 |
| Sonnet 5.5 | $0.00004 | $0.00139 |
| Haiku 4.5 | $0.00002 | $0.00070 |
Grade A, and why
crud 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 13d 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.
What it actually says
Create the CRUD domain $ARGUMENTS following the house pattern (express-prisma-pattern), plus the conventions in the prisma-schema-conventions skill.
Mandatory steps, in this order:
-
Detect the stack: check the project's
package.json.expresspresent → use theexpress-prisma-patternskill.@nestjs/corepresent instead → stop and tell the user Oficina doesn't maintain a NestJS pattern today (no active backend in the house stack runs NestJS); they can adaptexpress-prisma-pattern's principles by hand, or pull the retirednestjs-crud-patternskill from an older commit if they want the literal Nest structure back. Neither present → STOP and ask which pattern to follow; do not guess. -
Locate the project's data model document (UML/diagram in
docs/). If you can't find it or the entity isn't in it, STOP and ask for the fields — do not invent them. -
If a CRUD module/feature already exists in the repo, use it as the mold and keep the format identical.
-
Model the entity in
schema.prismafollowing the conventions (workspaceId, soft delete, UTC, and whichever others apply) and generate the migration with a descriptive name. -
Create the complete feature exactly as
express-prisma-patterndescribes — full folder structure, wiring, and validation approach. Controller → usecase → repository, wired by hand in a factory. -
Validate: project boots, migration applied, all 5 routes respond, lint and build clean.
-
Write the delivery log in
docs/log/(what was done, decisions, pending items), per the git-workflow rule. -
Self-verification (mandatory — do not skip): before declaring the domain done, check every item below and print the checklist with an explicit pass/fail per item. If any item fails, fix it before finishing; if it cannot be fixed, report it as a pending item — never declare done with a silent failure.
- Entity fields match the project's data model document (no invented fields)
- Migration applied with a descriptive name (table exists)
- Repository is an interface/abstract contract + Prisma implementation, never called directly by controller or business logic
-
workspaceIdcomes from the authenticated context; zero endpoints accept it in the body - Every repository query filters by
workspaceId - Input validated as the first thing that runs (zod, in whichever layer — controller or usecase — the project already validates at) — every non-negotiable rule of
express-prisma-patternfollowed - All 5 routes respond (POST, GET list, GET by id, PATCH, DELETE)
- Feature wired into the app: factory + route registration; lint and build clean
- Delivery log written in
docs/log/ - Any decision marked PENDING in the project's AGENTS.md was flagged, not silently decided
Non-negotiable rules: workspaceId from the authenticated context (never from the body), every query filters by tenant, permissions validated on the server, dates in UTC.
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.
- 13d ago Changed 3d85e6c4145d
- 1mo ago Changed · -10 tokens per session 7fa18b0c3012
- 1mo ago Changed · +1 lines · +9 tokens per session 86d103ce3457
- 1mo ago First seen · 30 lines · 20 tokens per session scan A e8c6342643f7
crud is a command published in the GitHub repository JoaoEquer/Oficina (2 stars, last pushed 13d ago), licensed MIT. It adds 19 tokens to every session and 697 once invoked, about $0.0001 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 2026-08-31.
Other commands, from other repositories
multi-backend
Run a backend-focused multi-model workflow for APIs, algorithms, data, and business logic.
dev-supabase
Configure and use Supabase as a backend (Auth, Database, Storage, Realtime, Edge Functions).
frappe-sql
Treat the task as Frappe-native SQL and ORM work.
ccc-multi-backend
Backend-focused multi-model workflow (research, ideation, plan, execute, optimize, review) with Codex as the backend authority. Use for API design, algorithms, database work and business logic.
saved-object
Generate a custom Saved Object type with registration, mappings, migrations, CRUD service, and management integration for a Kibana plugin.
init
Initialize configurations for Supabase local development.