chainsafe-zig-reviewer

Language-specific PR review for Zig code at ChainSafe (lodestar-z — consensus-critical Zig: SSZ, hashTreeRoot, Merkleization, hashing). Use this skill whenever the user is reviewing a Zig PR, asking for Zig code review, checking Zig for common mistakes, verifying a Zig diff against ChainSafe conventions, looking for allocator/leak issues, missing defer/errdefer, swallowed errors, undefined behavior, unjustified @setRuntimeSafety(false) or @ptrCast, integer-overflow bugs, or consensus-correctness gaps. EVEN IF the user does not explicitly ask for "Zig review" — triggers on "review this Zig PR", "is this Zig correct", "check this Zig function", "review the diff", "memory leak review", "undefined behavior check", "SSZ correctness", "allocator review". Findings emit at SOFT WARNING by default; memory-safety/UB and consensus-correctness paths promote to near-HARD-FAIL scrutiny. Do NOT use for Zig design (use chainsafe-zig-architect) or Zig implementation guidance (use chainsafe-zig-developer).

Ir para a instalação

Repositório: ChainSafe/engineering-handbook
Última atividade na origem: 17 de agosto de 2026 às 04:38
Idioma detectado do SKILL.md: inglês
Estrelas: 7
Forks: 1

Opções de instalação

Por padrão, está selecionado o prompt que primeiro revisa a origem. Você pode mudar para um comando direto ou baixar uma cópia local.

Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.

Um comando direto ignora o prompt de revisão. Verifique a origem antes de executá-lo.

npx skills add https://github.com/ChainSafe/engineering-handbook --skill chainsafe-zig-reviewer

O comando permanece em uma só linha. Role horizontalmente para revisá-lo antes de copiar.

Prefere uma cópia local? Baixe os arquivos disponíveis atualmente no SkillsMP.

Revise os arquivos de origem

Leia o SKILL.md e os arquivos complementares exibidos pelo SkillsMP antes de decidir se vai instalar.

SKILL.md

Instruções da origem · Visualização somente leitura

name: chainsafe-zig-reviewer
description: Language-specific PR review for Zig code at ChainSafe (lodestar-z — consensus-critical Zig: SSZ, hashTreeRoot, Merkleization, hashing). Use this skill whenever the user is reviewing a Zig PR, asking for Zig code review, checking Zig for common mistakes, verifying a Zig diff against ChainSafe conventions, looking for allocator/leak issues, missing defer/errdefer, swallowed errors, undefined behavior, unjustified @setRuntimeSafety(false) or @ptrCast, integer-overflow bugs, or consensus-correctness gaps. EVEN IF the user does not explicitly ask for "Zig review" — triggers on "review this Zig PR", "is this Zig correct", "check this Zig function", "review the diff", "memory leak review", "undefined behavior check", "SSZ correctness", "allocator review". Findings emit at SOFT WARNING by default; memory-safety/UB and consensus-correctness paths promote to near-HARD-FAIL scrutiny. Do NOT use for Zig design (use chainsafe-zig-architect) or Zig implementation guidance (use chainsafe-zig-developer).

# Zig Reviewer Reviewer skill for Zig PRs. Universal review framework at [`workflows/code-review.md`](../../workflows/code-review.md); full Zig-specific reference at [`languages/zig/reviewer.md`](../../languages/zig/reviewer.md). ## Severity tier **SOFT WARNING by default.** Two surfaces **promote to near-HARD-FAIL** — the reviewer stops and a human must engage: 1. **Memory-safety / undefined-behavior surface** — `@setRuntimeSafety(false)`, `@ptrCast` / `@alignCast`, raw many-item pointer arithmetic, `catch unreachable` on fallible ops, reliance on ReleaseFast semantics. 2. **Consensus-correctness surface** — SSZ encode/decode, `hashTreeRoot`, Merkleization. A byte wrong here is a fork, not a bug. Mirrors how Rust's `unsafe` is treated: the language allows it, the reviewer demands justification. ## Checklist ### Memory and allocators - Every allocation has a paired `defer` / `errdefer` free at the acquire. - Allocator is injected, not global. Ownership across returns and the C ABI is clear. - No leaks — `std.testing.allocator` stays green. Allocator choice fits the lifetime. ### Errors - No swallowed error unions — every `!T` is `try`-propagated, `catch`-handled, or exhaustively `switch`ed. - `errdefer` unwinds partial construction correctly. - `catch unreachable` only where success is a real invariant (near-HARD on fallible I/O). ### Undefined behavior and safety - `@setRuntimeSafety(false)` justified and scoped (unscoped/unexplained → near-HARD). - Integer-overflow intent explicit (`+%` / `+|` / `@addWithOverflow`); otherwise overflow impossible or handled. - `@ptrCast` / `@alignCast` sound — alignment honored, target layout guaranteed. No reading `undefined`; no reliance on ReleaseFast eliding checks. ### Comptime · C ABI - Comptime surfaces minimal and documented; no comptime where runtime fits. - Exported surface stable; cross-boundary memory ownership documented; null/optional handled at the boundary. ### Consensus correctness (lodestar-z and similar) - Spec-test coverage for SSZ / Merkleization / hashing changes, against the Ethereum consensus spec. - Fuzz coverage for serializers; determinism (endianness explicit, no floats on consensus paths). ### Tests, lint, format - Test blocks exist; `std.testing.allocator` used so leaks fail. `zig fmt --check` clean. ## Refusal The reviewer skill refuses to review and escalates if: - The diff contains non-Zig code. - The PR description is empty. - The diff touches code out of session scope. - **The diff introduces `@setRuntimeSafety(false)`, an unsafe pointer cast, or a consensus-path change without justification.** Refuse and escalate to the CODEOWNER. ## Phrasing - Lead with the concern: "This `@ptrCast` reinterprets `[]u8` as `[]u32`; the SAFETY note cites length but not alignment — what guarantees 4-byte alignment?" - Cite the spec for consensus findings. - `nit:` for taste-level. ## Related - Full reference: [`languages/zig/reviewer.md`](../../languages/zig/reviewer.md) - Gotchas: [`languages/zig/gotchas.md`](../../languages/zig/gotchas.md) - Universal review: [`workflows/code-review.md`](../../workflows/code-review.md) - Sister roles: `chainsafe-zig-architect`, `chainsafe-zig-developer`

Ver no GitHub

Voltar à pesquisa Voltar à instalação

chainsafe-code-review

Universal code-review framework at ChainSafe — language-agnostic principles plus the two reviewer modes (operator-reviewing-agent and agent-reviewing-PR). Use this skill whenever the user is reviewing a PR, asking for a code review, checking how to review, deciding what to flag vs fix, or handling a review disagreement. EVEN IF the user does not explicitly ask for "review" — triggers on "look at this PR", "check this diff", "what should I look for in this PR", "is this code OK", "how do I review", "approve this PR", "block this PR", "review disagreement", "review comments", "nit prefix", "should I flag this". For language-specific review (Go, Rust, TypeScript, Solidity), additionally load the corresponding language reviewer skill (chainsafe-<lang>-reviewer). For PR authoring (outgoing), use chainsafe-pr-author.

2026-08-17

chainsafe-pr-author

How to open and shepherd a pull request at ChainSafe — language-agnostic PR authoring guidance. Use this skill whenever the user is opening a PR, drafting a PR description, deciding how to scope a PR, handling reviewer comments, structuring commits within a PR, declaring AI-generated content, or shepherding a PR through review. EVEN IF the user does not explicitly say "PR" — triggers on "open a pull request", "draft a PR description", "what should I put in the PR", "PR is too big", "split this PR", "how do I respond to this reviewer comment", "self-resolve threads", "AI declaration in PR", "scope drift", "commit message", "stack of PRs". For the actual code-change workflow that produces the PR, use chainsafe-research-plan-implement. For PR review (incoming), use chainsafe-code-review.

2026-08-17

chainsafe-research-plan-implement

Research-first coding workflow that gates ALL non-trivial code changes behind a human-approved written plan. Use this skill whenever the user asks for any substantive code change — implementing a feature, refactoring, fixing a multi-file bug, adding an endpoint, integrating a service, migrating systems, or changing how an existing system works — EVEN IF they do not explicitly ask for "research" or a "plan." It enforces three artifacts (research.md, plan.md, annotated plan) that gate implementation behind operator approval. Triggers on "implement", "refactor", "fix this bug", "add a feature", "integrate", "migrate from X to Y", "rewrite", "wire up", "extend", "I want to add". Also covers breaking planned work into tracked units — "break this into issues", "decompose this", "create the epic", "what milestones do we need", "is this issue too big". Do NOT use for one-line typo fixes, pure formatting, lint-only changes, or other trivial single-file edits.

2026-08-17

Ver todas deste repositório

chainsafe-zig-reviewer

Informações da origem

Opções de instalação

Revise os arquivos de origem

Mais deste repositório

Mais deste repositório