Desenvolvimento

Esta skill de limpeza automática é ativada antes do término da sessão, utilizando quatro subagentes paralelos para detectar atualizações de documentos, padrões repetitivos, pontos de aprendizado e tarefas de acompanhamento, enquanto um subagente de verificação remove duplicatas e apresenta opções ao usuário. Gatilhos incluem /session-wrap, finalização de sessão, limpeza de sessão e finalização de tarefa.

Use when simulating a technical interview to build confidence, expose reasoning gaps, and train structured communication under pressure.

Create GitHub PR from issue number ∨ objective. Triggers when user mentions GitHub PRs, pull requests, opening PR. Phrasings: "open a PR", "create a pull request", "PR for issue #N".

Create, file, open, log, ∨ track GitHub issue. Always-gated by `core:socratic` skill — concrete intent passes ≤ 1 turn, vague intent triggers dialogue until convergence. Triggers when user mentions GitHub issues, bug reports, feature requests, ∨ wants to track work on GitHub. Phrasings: "file an issue", "open a bug", "track this on GitHub".

Grades and improves CLAUDE.md (Claude Code) and AGENTS.md (Codex/OpenCode) configuration files. Use when asked to grade, score, evaluate, audit, review, improve, fix, optimize, or refactor agent config files. Triggers on 'grade my CLAUDE.md', 'score my AGENTS.md', 'is my CLAUDE.md too big', 'improve my agent config', 'fix my CLAUDE.md', 'optimize context usage', 'reduce tokens in CLAUDE.md', or 'a

Read-only drift detector. Diffs SPEC.md vs current code ∧ reports violations grouped by severity. ⊥ writes — suggests remedies via spec ∨ build skills, ⊥ invokes them. Triggers when user asks to check drift, audit spec, verify invariants, ∨ ask if code still matches spec. Phrasings: "check drift", "audit the spec", "does the code still match §V", "check invariants", "spec vs code", "is the spec st

Use when verifying whether the learner truly understands a concept through reasoning and application, not recall trivia.

When the user wants a thorough code review of changes, a PR, a file, or a feature branch. Use when the user says "review this," "code review," "check my code," "what's wrong with this," "review my PR," "look over this," "critique this," "review before merge," or when submitting code for quality assessment. Performs multi-pass review covering correctness, security, performance, maintainability, and

Avalia modelos de geração de código em HumanEval, MBPP, MultiPL-E e mais de 15 benchmarks com métricas pass@k. Este padrão da indústria do Projeto BigCode, usado pelos leaderboards do HuggingFace, é ideal para comparar modelos de código e medir a qualidade da geração de código.

When you or the user are going in circles, stuck on a problem, or repeating failed approaches. Use when the user says "I'm stuck," "this isn't working," "we keep going in circles," "try something different," "nothing works," "help," or when you detect that you've attempted the same category of solution 2+ times without success. Forces a structured step-back, reframe, and fresh approach.

Cut SemVer release for single-package repo — bump `version` in detected manifest, commit, create annotated tag `v<x.y.z>`. Local only, ⊥ push, ⊥ GitHub Actions. Triggers when user says "release", "tag", "ship", "cut a version", "bump version".

Enforces characterize-before-change discipline with continuous green tests, separating structural changes from behavioral ones.