didacrios/ai-coding-protocols

Modular AI coding protocols based on Recursive Language Modeling (RLM). Standardizing codebase exploration, recursive dependency tracing, and incremental development for Cursor, Claude, and Gemini.

2Stars on the repository
10Mods indexed here, across every type
5d agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown

code-review

01

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Review a branch, commit, or pull request with emoji-tagged findings that separate blocking bugs from advisory notes, each blocking comment carrying a concrete fix. Use when reviewing a diff, auditing a PR, or giving structured feedback on code changes. Don't use it to review your own work in the session that produced…

not rated 2 changed 20d ago A 66 tokens original MIT

documentation

02

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Write in-code comments and system documentation that explain why rather than what, using the Diátaxis framework for external docs. Use when adding TSDoc or PHPDoc, creating a README, tutorial, how-to guide, explanation, or reference page, or documenting an API or schema.

not rated 2 changed 20d ago A 59 tokens original MIT

php-standard

03

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Apply opinionated PHP 8.2+ and PHPUnit standards — declare(stricttypes=1) everywhere, PSR-12 and PSR-4, value objects, and a test-first loop with Given-When-Then naming. Use when writing, reviewing, or testing PHP, wiring Composer autoloading, or setting up PHPUnit.

not rated 2 changed 20d ago A 72 tokens original MIT

plan-initiative

04

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Shape a rough engineering initiative into one scoped, outcome-oriented brief through strict-stop discovery, explicit in-scope and out-of-scope boundaries, and named risks and assumptions. Use when an Engineering Manager asks to plan, frame, or structure an initiative, when scope or ownership is still fuzzy, or when…

not rated 2 changed 20d ago A 76 tokens original MIT

project-bootstrap

05

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Bootstrap a project for these skills by scanning the codebase to detect stack, architecture, and conventions, then generating one project-overview skill and tool-specific wiring. Use on the first run in a new project, when a skills directory exists without project-specific skills, or when the user asks to initialize…

not rated 2 changed 20d ago A 67 tokens original MIT

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Explore unfamiliar code before changing it by tracing consumers and dependencies recursively, then landing one verified step at a time. Use when starting a task in code you have not read, tracing how a change propagates, planning edits across multiple files, or debugging by following execution paths. Don't use it when…

not rated 2 changed 5d ago A 76 tokens original MIT

refactor

07

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Refactor code surgically without changing behavior — trace consumers first, require a green test suite, then transform one method per step with tests after each. Use when cleaning up after a feature, reducing method length or complexity, extracting helpers, aligning code with a stack standard, or renaming symbols to…

not rated 2 changed 20d ago A 65 tokens original MIT

skill-generator

08

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Create or update a project-specific skill when a convention, module, or integration recurs across the codebase. Use when a new module introduces conventions no existing skill covers, when the same pattern appears in two or more places, when a new library or framework is adopted, or when the user asks for a project…

not rated 2 changed 20d ago A 67 tokens original MIT

typescript-standard

09

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Apply opinionated TypeScript and testing standards — strict mode with no any, 10-20 line methods, and a test-first loop using data builders, Faker, and Given-When-Then. Use when writing, reviewing, or testing TypeScript, setting up Jest or Playwright, or checking TS code against a project's conventions.

not rated 2 changed 20d ago A 71 tokens original MIT

plan-diagram

10

didacrios/ai-coding-protocols

Skill Claude CodeCodex

Draw a coarse boxes-and-arrows model (C4-style) inside a plan, spec, or issue so the work is visible first. Use when planning, specifying, writing issues, or asked to diagram or sketch a change; not for call sequences.

not rated 2 5d ago A 55 tokens original MIT

At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: