v0.1.4 — Deeper Rust learning and Japanese writing guidance

Rust Learning Lab now explains more of the reasoning behind Rust notation and connects type-level evidence to implementation behavior, system boundaries, and focused learning support.

• Add a symbol inventory and executable contrasts covering borrowing, lifetimes, dereferencing, paths, generic arguments, closures, errors, macros, and recursion. Explain what each symbol enables and how its role changes with context.
• Deepen Rust 2024 guidance with edition-specific examples for patterns, opaque return-type capture, temporary destruction, macro matching, and boxed-slice iteration. Distinguish edition support from feature stabilization and minimum-toolchain verification.
• Add guidance for reading signatures, validating domain types, and reasoning about retries, cancellation, persistence, and compatibility. Separate what Rust types establish from what needs implementation, protocol, or runtime evidence.
• Refine learning support and add Japanese writing rules with optional Suiko recommendations. Preserve technical conditions and uncertainty, avoid repeated advice, and require no Python. Update English and Japanese guidance and both plugin manifests to 0.1.4.

Validation: CI passed, all 78 Rust doctests passed locally, and bilingual coverage, Rust example parity, local Markdown links, and distribution metadata passed validation. The new Japanese writing reference was checked with Suiko 0.3.10. Both Codex and Claude Code installed version 0.1.4 from the published marketplace; distributed plugin contents matched the repository, excluding Claude Code's runtime marker.

Conversational evaluation scenarios were expanded but have not been executed. The examples were tested with Rust 1.98.1; this does not establish Rust 1.85 compatibility or measured learning effectiveness.

Start a new Codex conversation or restart Claude Code to load the updated instructions.

Full changes: v0.1.3...v0.1.4

v0.1.3 — PR and commit learning reviews

Create educational Rust review documents from PR URLs or numbers, commits, and commit ranges. The skill records the compared revisions and explains behavior, syntax, argument order, ownership, and design decisions from the matching source.

• Store separate reviews under docs/rust-learning-review/ and update English and Japanese README indexes after each document creation or update.
• Keep generated reviews and indexes out of Git and routine searches with .gitignore, .rgignore, and shared agent guidance. Read only the requested material or files needed for review maintenance; do not automatically copy learning notes into agent memory or repository rules.
• Update bilingual instructions, reference guidance, README examples, and evaluation scenarios. Preserve existing user notes and distinguish historical source inspection from tests actually run.

Validation: CI passed, all 14 Rust doctests passed, local Markdown links and both host manifests passed validation, and Git/ripgrep exclusions were checked. Codex and Claude Code installed version 0.1.3 from the published marketplace, with installed content matching the plugin source. New conversational evaluation scenarios are documented but have not been run; ignore rules and guidance are not an access-control boundary.

Start a new Codex conversation or restart Claude Code to use the updated instructions.

Rust Learning Lab v0.1.2

All eight learning and review references now have English and Japanese editions, with reciprocal links and same-language cross-references. The skill instructions and evaluation scenarios are bilingual too.

The skill chooses references according to the language of the current explanation or document and avoids loading both translations unnecessarily. Code, numerical research findings, and limitations are preserved across versions.

Repository guidance now records bilingual maintenance, review-document behavior, verification, and publication practices. Claude Code imports the same guidance.

Validation: 14 doctests passed across both languages, matching Rust code blocks and paired files checked, local links and fragments checked with no errors, and host manifests validated. These checks do not establish model behavior or learning effectiveness.

See English documentation or 日本語の説明.

Rust Learning Lab v0.1.1

Rust Learning Lab helps you understand and review Rust code through syntax, design rationale, types, ownership, and concrete verification.

• Shared learn-rust skill for Codex and Claude Code, with installable marketplaces for both hosts.
• Focused practice and debugging guidance informed by programming education and AI assistance research, with explicit limitations.
• Code reviews explain each finding's location, conditions, impact, and fix; detailed learning reviews can produce reusable Markdown documents.
• Responses adapt to the conversation's language; English and Japanese READMEs link to each other.

Validation: both host manifests validated, seven Rust documentation examples passed (including an intentional compile failure), and distribution metadata checked in GitHub Actions. This does not establish model review accuracy or learning effectiveness.

See English setup or 日本語の使い方.