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 filipebraida/adonisjs-starter-kit --skill crudgit clone --depth 1 https://github.com/filipebraida/adonisjs-starter-kitWrote 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/filipebraida/adonisjs-starter-kit/crud)<a href="https://agentmods.dev/skills/filipebraida/adonisjs-starter-kit/crud"><img src="https://agentmods.dev/badge/skills/filipebraida/adonisjs-starter-kit/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/skills/filipebraida/adonisjs-starter-kit/crud"><img src="https://agentmods.dev/badge/skills/filipebraida/adonisjs-starter-kit/crud.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00066 | $0.02765 |
| Opus 5.5 | $0.00026 | $0.01106 |
| Sonnet 5.5 | $0.00013 | $0.00553 |
| Haiku 4.5 | $0.00007 | $0.00277 |
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 2mo 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.
How it starts
The opening of the file, as written. The whole thing — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CRUD
A CRUD feature is a stack, not a single file. The route declares one router.resource(...) scoped by auth middleware; the controller has 5–7 tiny methods that only orchestrate; each mutation has a VineJS validator; a BasePolicy gates every method; each write goes through a single-purpose action class; a transformer produces per-page shape variants; the read side lives in queries/; the UI is Inertia pages under <mod>/ui/pages/. The point is that no single layer holds more than its own responsibility, and adding the next resource is copy-shape, not think-shape.
Rules
Every HTTP method — including index — follows the same 3-step spine:
- Authorize via
bouncer.with(XPolicy).authorize(...). Awhere('owner_id', user.id)clause in the query is not a substitute — the policy is the source of truth. - Validate via
request.validateUsing(...). - Render / redirect — either
inertia.render('...', props)for reads orresponse.redirect().back()/.toRoute(...)for writes. Writes call an action in between.
Authentication is the middleware's job — controllers never call authenticate(); use auth.getUserOrFail() when the handler needs the user. Anything past .authorize() that also does an owner check is redundant; anything before it that touches the DB other than loading the resource being authorized is a leak.
Layer conventions
- Route —
router.resource('/entities', EntitiesController).only([...]). Guard by wrapping in arouter.group(() => {...}).middleware(middleware.auth())block (see [[routes]]). Custom verbs go as separaterouter.post(...)calls, not extra controller methods. - Controller (thin) — one method per resource action. Body order:
policy → validator → action call → render/redirect. No business logic, no queries longer than one line, no ORM calls beyondModel.findOrFail. - Validator — VineJS
vine.create({...})per mutation, exported from<mod>/validators/<entity>.ts(one file per entity —validators/users.ts,validators/tokens.ts). Usevine.withMetaData<{...}>().create({...})when a rule depends on route params (e.g. unique-except-self on edit).vine.compile()is deprecated — usevine.create(). - Policy —
BasePolicysubclass at<mod>/policies/<entity>_policy.ts. One method per gated action. Returnboolean | AuthorizerResponse. Gate before validating:await bouncer.with(XPolicy).authorize('action', resource?). See [[authorization]]. - Action — one class per write in
<mod>/actions/<verb_entity>.ts.Inputis a plain interface; the class has oneasync handle(input)method. Never takesHttpContext. Side effects go through domain events — see [[actions-events]]. - Transformer —
BaseTransformer<Model>withtoObject()as the base + variants (forList,forEdit,forSharedProps,forProfile). CallTransformer.transform(model).useVariant('forEdit')orTransformer.paginate(rows, meta).useVariant('forList'). Never send raw Lucid instances to Inertia — the shape leaks and the response type drifts. - Query — read-only,
<mod>/queries/list_<entities>.ts. Full patterns (list queries + read models for aggregate screens) live in [[queries]]. - Inertia pages — under
<mod>/ui/pages/<entity>/. Index is a full page; create/edit are modals mounted over the index. See [[inertia]].
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.
- 2mo ago First seen · 217 lines · 66 tokens per session scan A 64b1a08b96e6
crud is a skill published in the GitHub repository filipebraida/adonisjs-starter-kit (96 stars, last pushed 2mo ago), licensed MIT. It adds 66 tokens to every session and 2,765 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 2026-08-30.
Other skills, from other repositories
supermemory
Supermemory is a state-of-the-art memory and context infrastructure for AI agents. Use this skill when building applications that need persistent memory, user personalization, long-term context retention, or semantic search across knowledge bases. It provides Memory API for learned user context, User Profiles for…
tuturuuu-external-apps
Connect branded external apps to Tuturuuu authentication, workspace administration, storage, and content APIs.
swarmauri-add-base-class
Add or update a Swarmauri SDK base class in pkgs/base/swarmauribase. Use when Codex needs to implement a base class over a core interface, update ResourceTypes, wire InterfaceRegistry, set ComponentBase registration, exports, and focused tests.
ui-beats
Use this skill when users want to add, customize, or troubleshoot UI Beats components in React/Next.js projects. It covers component selection, shadcn registry installation from uibeats.com, the UI Beats MCP server, motion and reduced-motion handling, and integration patterns for animated sections.
tuturuuu-cli
Install, authenticate, troubleshoot, or publish the Tuturuuu CLI and SDK.
tuturuuu-realtime
Implement and operate Tuturuuu realtime rooms, collaborative programming, presence, cursors, durable state and private checkpoints.