Ai Olympus Laravel Package

Agents

Behavioral guidelines to reduce common LLM coding mistakes. Merge with project-specific instructions as needed.

claude-code

cursor

analyze-problem

Use when structured problem analysis for debugging, root cause identification, and breaking down complex issues before proposing solutions

claude-code

analyze-support-issue

Use when a JIRA issue from a support team needs an analysis a non-technical colleague can act on. Decides whether it is an application bug, a known bug, correct behaviour the user did not expect (with or without a documentation article), a documentation error, a missing feature, a data request, or something outside the application, and publishes one plain-language TL;DR comment in the language of the issue. Runs unattended; every statement in the comment is verified in this run and nothing is estimated.

claude-code

api-review

Use when reviewing HTTP API design in a PR or change set — endpoints, routes, HTTP methods, status codes, idempotency, and input validation. Treats the API as a consumer-facing contract and flags resource-orientation, method-semantics, status-code, and trust-boundary violations. Read-only.

claude-code

assignment-compliance-check

Use when checking that the pull request implementation actually fulfills the business requirements stated in the linked issue or task. Returns a plain-language Assignment Compliance markdown block on every run that has a linked tracker, always carrying one of three verdicts: every acceptance criterion is met, a named criterion is not met and what is missing, or the assignment states no explicit criteria plus the basis the state was judged on. Only a run with no linked tracker returns a skip status. The block carries no file paths, line numbers, or code snippets, and it is embedded in the consolidated tracker comment, never in the GitHub PR comment. No local file is created.

claude-code

auto-interactive-testing

Use when every open GitHub issue labelled interactive-testing must be worked through as one unattended batch. Fixes and merges the defects the issues already link, runs each issue's scenarios with the interactive-testing skill on the freshly pulled default branch, resolves and merges every new defect it finds, re-tests, cleans up its test data, then posts the verdicts on each issue and closes it when every scenario passed.

claude-code

blast-radius-analysis

Use when a planned code change must be mapped to everything that depends on it before implementation starts. Produces a read-only, evidence-graded impact report — affected components, risks, unknowns, a targeted verification plan, and implementation constraints — at a depth set by the FAST / STANDARD / CRITICAL tier.

claude-code

class-refactoring

Use when refactor PHP classes to improve structure, readability, and maintainability while preserving behavior

claude-code

cleanup-local-branches

Use when cleaning up local Git branches after origin pruning. Deletes local branches whose upstream was deleted on origin (marked gone) and local branches with no origin counterpart that have been inactive for more than six months, while always protecting the current branch and the default branches. Previews every deletion before running it.

claude-code

code-review

Use when senior PHP code review focused on architecture, business logic, and risk detection. Read-only.

claude-code

code-review-bugsnag

Use when run code review for a Bugsnag error and publish results to the linked GitHub PR and the Bugsnag error

claude-code

code-review-github

Use when perform code review for GitHub pull requests and post findings as PR comments plus a non-technical summary to every linked issue

claude-code

code-review-jira

Use when run code review for JIRA issues and publish results to GitHub PR and JIRA

claude-code

compact-project-memory

Use when docs/memory/PROJECT_MEMORY.md was just written to — a new entry appended or an existing one edited — and the run that wrote it is about to report completion. Shrinks the token footprint of only the entries the write actually touched, plus at most 3 demonstrably related ones, without ever losing a recorded fact; a true no-op when the file carries no diff.

claude-code

create-issue

Use when create a single issue from provided text without modifying its content

claude-code

create-issues-from-text

Use when break down assignment into multiple structured issues

claude-code

create-missing-tests-in-pr

Use when a PR review already exists and its missing tests must be completed. Reads the pull request code review, verifies which recommended behaviours are already protected, adds only the tests for a missing behavioural contract using the create-test skill, and verifies 100% coverage of the changed code.

claude-code

create-test

Use when creating or updating tests for current changes. Writes the minimum tests that protect the changed behaviour, risk, or contract at its owner boundary, then verifies full coverage of the changed code.

claude-code

deliver-page-redesign

Use when one page of a running application must be redesigned end to end from its URL — analysed, proposed with a preview of every state, refined until the user approves the previews, then implemented with the existing design system, reviewed, and verified in a real interactive browser before the pull request is reported.

claude-code

design-system

Use when generating, auditing, or reviewing the visual design system of a Laravel app — Tailwind tokens, Filament theming, Blade/Livewire component consistency, and visual-polish audits.

claude-code

diagram-design

Use when a change, analysis, or document needs a diagram — architecture, request flow, sequence, state machine, ER model, timeline, or comparison — and a picture explains it better than prose. Produces one self-contained HTML + inline SVG file with no build step, no external asset, and no runtime dependency, styled by a fixed editorial design system and validated against a density, contrast, and geometry checklist before it ships.

claude-code

docker-patterns

Use when writing or reviewing Docker and docker-compose setups for a Laravel application — multi-stage PHP-FPM images, services (nginx, MySQL, Redis, queue worker, scheduler, Vite build), healthchecks, secrets, and image hardening.

claude-code

e2e-testing

Use when writing or stabilizing Playwright end-to-end browser tests against a Laravel app — but only when the project already has Playwright; otherwise defer to manual testing or Pest/Dusk.

claude-code

frontend-a11y

Use when building or reviewing accessible UI in a Laravel app — semantic Blade markup, accessible forms, keyboard navigation with Alpine, focus and live-region management for Livewire updates, contrast, and Filament accessibility.

claude-code

frontend-design-direction

Use when the work is not just making UI function but making it feel purposeful and polished — choosing a deliberate design direction for a Laravel/Blade/Livewire/Filament interface.

claude-code

frontend-patterns

Use when building Livewire/Blade/Alpine UI in a Laravel app — component composition, state placement, performance, forms, and loading/empty/error states.

claude-code

frontend-slides

Use when building standalone HTML/CSS/JS presentation slide decks — self-contained single-file decks with viewport-fit layout, keyboard navigation, and browser Print-to-PDF export.

claude-code

github-issue-triage

Use when GitHub issues must be prioritized, sorted, or labelled by type, when the open backlog must be swept — issues already merged closed, epics kept flat, issues re-parented under the epic that owns them — or when the issues labelled `analyze` must be analysed and the analysis published on them. Seeds the priority and type label taxonomy and assigns the derived priority to every open issue, so work can be filtered and picked in priority order.

claude-code

github-release-roadmap

Use when planning a GitHub release roadmap for one repository — enumerates issues, labels, milestones, releases, and Projects with `gh`, proposes a versioned release scope for explicit approval, then applies the confirmed plan idempotently without publishing a release or pushing code.

claude-code

git-workflow

Use when choosing a Git branching strategy or handling merge vs rebase, conflicts, undoing mistakes, and release tagging — complementing the commit/PR conventions in the git rules.

claude-code

interactive-testing

Use when verifying user-visible project changes against an assignment and the actual code diff. Requires the testing agent's own interactive browser, checks security risks, verifies suspected vulnerabilities, and drives confirmed threats through repair and re-testing.

claude-code

laravel-authorization-review

Use when reviewing authorization / access control in a Laravel project — find IDOR / broken object-level authorization (BOLA), audit which routes are unprotected, check policy / gate coverage, or sanity-check a new endpoint in a PR. Walks the authorization chain of every HTTP route (middleware → authorize/policy/gate → query scoping → API Resource output), anchors every finding to real `php artisan route:list --json` output plus a cited `file:line`, classifies by evidence, and produces a per-route coverage map. Read-only / advise-only — never edits code.

claude-code

laravel-security

Use when building, configuring, or hardening security-sensitive Laravel features — authentication, authorization, Eloquent safety, CSRF/XSS, API security, file uploads, secrets, and production configuration. Provides condensed, copy-ready secure defaults for Laravel 12/13 / PHP 8.4/8.5.

claude-code

laravel-telescope

Use when analyzing Laravel Telescope requests from URL and DB. Loads Telescope entries, matches the same request in database tables, and proposes practical optimizations.

claude-code

latency-critical-systems

Use when working on latency-sensitive Laravel paths — realtime dashboards, streaming, queues, caches, or execution gateways — where p95 latency and data freshness matter.

claude-code

machine-payments-protocol

Use when implementing, designing, or reviewing the Machine Payments Protocol (MPP) HTTP 402 payment flow in a Laravel/PHP application — middleware placement, payment verification, provider abstraction, pricing, and MCP tool billing. Separates protocol-normative behavior (cited from the IETF draft and its reference implementations) from illustrative naming and generic API-security hardening.

claude-code

merge-github-pr

Use when safely merge GitHub pull requests that are ready

claude-code

mysql-patterns

Use when designing MySQL schema features or applying advanced MySQL patterns in Laravel — upserts, JSON columns, full-text search, partitioning, replication/read-write splitting, and deadlock handling — beyond the query tuning already in the SQL rules.

claude-code

mysql-problem-solver

Use when analyze real MySQL query and schema problems using code inspection, schema review, and EXPLAIN when available

claude-code

page-redesign

Use when an existing application page must be redesigned so its element layout is logical, intuitive, and readable for operators who are not IT people — warehouse, workshop, and shop-floor staff who use the software to process orders as fast as possible. Produces a grounded redesign proposal, a rendered preview screenshot of every state the main view hides, and an implementation-ready handoff a developer can build from without asking a follow-up question. Never changes the application's main layout shell unless the caller explicitly orders it.

claude-code

postgres-patterns

Use when designing PostgreSQL schema features or applying advanced Postgres patterns in Laravel — GIN/BRIN/partial/covering indexes, jsonb, ON CONFLICT upserts, SKIP LOCKED queue workers, cursor pagination, RLS, and timestamptz/numeric typing — beyond the query tuning already in the SQL rules.

claude-code

prepare-issue-context

Use when preparing data and context before /resolve-issue, TDD, or CR runs. Loads the assignment, extracts every concrete user scenario from the task description and acceptance criteria, maps each scenario to the codebase, seeds the development database with the records needed to reproduce the bug or feature end-to-end, and reports any gap that would force the implementing agent to hallucinate.

claude-code

process-code-review

Use when processing pull request code review feedback. Finds the latest PR for a task, resolves review comments, updates review status, and triggers the next review cycle.

claude-code

product-capability

Use when a PRD or product intent is clear but the implementation constraints are not — turns a vague capability ask into an engineering-ready plan that exposes invariants, interfaces, and unresolved decisions before any code is written.

claude-code

pr-summary

Use when summarizing current PR changes for the development and product team. Analyzes all commits in the current branch, explains what broke and what the fix changes, and produces a human-readable report that can be posted as a GitHub PR comment (Markdown), a JIRA comment (ADF), or a Bugsnag comment (plain text).

claude-code

redis-patterns

Use when using Redis in a Laravel app — caching strategies, atomic/distributed locks, rate limiting, stampede protection, pub/sub, pipelines, and key/TTL design beyond raw query tuning.

claude-code

refactor-entry-point-to-action

Use when refactoring controller, job, command, listener, or Livewire entry-point logic into a dedicated Action class while preserving behavior and response contracts.

claude-code

resolve-issue

Use when resolving an issue from any supported tracker (GitHub, JIRA, Bugsnag). Detects the source automatically from the provided link or ID, implements a safe fix or feature, validates with tests, and creates a pull request.

claude-code

rewrite-tests-pest

Use when rewriting existing tests to Pest syntax. Preserve behavior, follow project testing conventions, reduce duplication where helpful, and verify rewritten tests are deterministic and passing.

claude-code

security-audit

Use when the user explicitly asks for a security audit of the whole repository or of one subsystem, and in Diff mode over the current diff of every code review. Runs a budgeted, reproducible, coverage-driven audit in the quick, standard, or deep profile: a coverage inventory before hunting, read-only hunters in waves, a fresh coverage critic after each wave, and a new independent verifier for every candidate. Writes metadata.json, coverage-ledger.json, findings.json, and REPORT.md, and validates the JSON against its schemas. An incremental run compares an earlier run with the git diff and reuses its coverage only where the bytes of the security path are unchanged.

claude-code

security-bounty-hunter

Use when hunting for exploitable, remotely reachable vulnerabilities in a PHP/Laravel codebase for responsible disclosure or a bounty submission, not a general best-practices review. Biases toward user-controlled attack paths that pay and discards low-signal noise.

claude-code

security-review

Use when performing a focused security review for Laravel/PHP projects. Prioritize real exploitability, business logic flaws, and high-risk vulnerabilities.

claude-code

security-threat-analysis

Use when analyzing a specific security threat from a referenced source (CVE, GHSA, security advisory, blog post, or write-up). Produces a human-readable remediation report with step-by-step instructions an AI agent can follow to eliminate the threat in the current project.

claude-code

seo

Use when auditing, planning, or implementing SEO in a Laravel app — crawlability, indexability, JSON-LD structured data in Blade, Core Web Vitals, on-page tags, keyword mapping, competitor gap analysis, E-E-A-T content quality, and measurement.

claude-code

simplification-audit

Use when the user explicitly asks for an audit of the codebase or to refactor a part of the codebase. Coordinates a bounded, read-only pass over every subsystem in scope and returns ranked simplification recommendations for data structures, state representation, control flow, algorithms, and ownership, without changing a single file.

claude-code

skill-creator

Use when creating a new Agent skill in this repository. Generates a SKILL.md that follows project conventions, passes skill-check validation, and updates the changelog and readme.

claude-code

smartest-project-addition

Use when you want exactly one high-impact, concrete proposal for the next project addition.

claude-code

test-assignment

Use when a task's pull request must be tested against its assignment and the test must prove that nothing else broke. Maps every input that reaches the changed behaviour, builds a test matrix of real data, runs the whole test suite and CI on the final head, and exercises the running application against the base branch. Report mode changes nothing and publishes a test report; fix mode adds the missing tests and fixes, then prepares the pull request for merge without merging.

claude-code

test-audit

Use when a test is about to be written or changed, when a diff that touches tests is reviewed, or when tests must be audited for low value. Holds the single test-value bar — behaviour, risk, or contract first, coverage only as verification — with a lightweight authoring gate, a read-only code-review lens, a diff mode the implementer runs before the pull request to delete tests that prove no assignment logic while keeping 100% coverage, and an explicit read-only audit of a named scope that proposes KEEP / CONSOLIDATE / MOVE / DELETE with evidence.

claude-code

test-driven-development

Use when implementing a feature or bugfix with strict TDD. Enforce failing-test-first, minimal implementation, and safe refactoring.

claude-code

tester-cookbook

Use when preparing a concise QA report for an internal tester from a JIRA task and its linked pull requests — focused on what the tester should report back to the dev team — and posting it as a JIRA comment.

claude-code

understand-propose-implement-verify

Use when following a strict problem-solving loop: understand, propose, implement, verify.

claude-code

verify-merge-readiness

Use when a GitHub issue, a GitHub pull request, or a JIRA issue must have its pull request brought to a merge-ready state without merging, then summarized in one source-issue TL;DR while superseded preparation comments are removed safely.

claude-code

vite-patterns

Use when configuring or optimizing Vite (laravel-vite-plugin) asset bundling in a Laravel app — entrypoints, the @vite Blade directive, HMR, env vars, aliases, manifests, code splitting, and production builds.

claude-code

web-article-writer

Use when writing, rewriting, or adapting a publication-ready article for a website or blog. Matches the site's editorial conventions, verifies factual claims, applies restrained on-page SEO, and can prepare an optional illustration brief.

claude-code