cross-harness-skills

Review a PR through the CodeReviewCompleteness persona. Use when asked to check whether a PR fully addresses the objectives it claims to address in its title, description, cited issue, PRD, epic, acceptance criteria, or linked discussion.

Review a PR through the CodeReviewCorrectness persona. Use when asked to check whether the implementation does the job correctly, avoids bugs and regressions, preserves data and API contracts, handles relevant edge cases, and has tests that would catch the important failures.

Review a PR through the CodeReviewElegance persona. Use when asked to review code for minimal design, architectural leverage, elegant abstractions, simplification, removing unnecessary moving parts, or finding the design move that sidesteps an entire class of problems.

Review a PR through the CodeReviewIdomatic persona. Use when asked to check whether code is idiomatic for the implementation language, follows standard library and ecosystem conventions, avoids unnecessarily novel implementations, and matches established patterns in this repository.

Review a PR through the CodeReviewScope persona. Use when asked to check whether a PR stays within its intended scope, avoids crossing into unrelated topics, separates incidental cleanup from behavior changes, and keeps review lanes clear.

Review a PR through the CodeReviewSecurityReview persona. Use when asked to check whether code is safe, avoids backdoors, vulnerable dependencies, injection attacks, auth bypasses, secret leaks, unsafe file/network behavior, and other security regressions.

Coordinate a delegated full multi-persona code review for a PR or local change. Use when asked to review a PR, run review agents, get code review feedback, or iterate on review feedback using all CodeReview personas: elegance, idiom, correctness, completeness, security, YAGNI, and scope. Invoking this skill is an explicit request to launch reviewer subagents.

Review a PR through the CodeReviewYAGNI persona. Use when asked to check whether a PR adds unnecessary implementation surface beyond the stated objective. First verify the issue or PR acceptance criteria and current repository direction; do not flag behavior that is explicitly required, already established as a repo pattern, or needed for plausible production states.

Audit and improve a code-review persona or review-coordinator skill using historical agent sessions. Use when asked to evaluate whether a review agent is useful, noisy, overreaching, missing meaningful findings, or to update a review skill or agent definition based on observed Moraine/session-history behavior.

Create ready-for-review pull requests on GitHub. Use when the branch is ready to open a PR; when the user asks to create, push, or open a PR; after local changes are committed; or when documenting user-facing features, bug fixes, performance optimizations, validation, changelog impact, or operational impact in a PR that follows repository conventions without AI attribution.

Turn a rough feature, bug, refactor, architecture, documentation, or operations idea into a ready-to-implement local plan file. Use when the user asks to crystallize, shape, refine, plan, spec, or de-risk ambiguous work before implementation; the workflow researches the codebase, public prior art, and relevant history, debates the proposed approach with subagents, and writes a final plan with no open questions.

Cut and publish repository releases from a version argument such as /release X.Y.Z or a request to use $cross-harness-skills:release. Use when Codex is asked to run a release process, bump release-managed versions, create and merge a release PR, push a vX.Y.Z tag, write GitHub release notes for users, verify release automation, and confirm published artifacts or package registry entries. Start or continue a Codex goal for the release.

Start repository development work safely. Use when an agent takes on new code, docs, test, CI, packaging, or release work; when it must create or verify an isolated branch/worktree; when sibling agent collisions are possible; or before editing files for a contributor task.

Use when creating, reviewing, or adapting an agent skill that should remain portable across Codex, Claude Code, and other compatible harnesses.