elijj

@elijj source repo

17 skills
0 followers
13 hours ago last updated

▌
Prp Pr · elijj bundle
Use this skill when the user invokes `/prp-pr`, asks to create a GitHub pull request from the current branch, wants commits analyzed into a PR title/body, needs PR templates discovered, wants the branch pushed with upstream tracking, or wants PRP plan/report artifacts referenced in a PR. It validates git state, uses GitHub CLI when available, and recommends `prp-commit` or `code-review` as needed.

0 installs
▌
Prp Plan · elijj bundle
Use this skill when the user invokes `/prp-plan`, asks for a PRP, wants an artifact-producing implementation plan, provides a PRD or feature description that needs codebase analysis, pattern extraction, validation commands, acceptance criteria, and a self-contained plan file. Use this for deeper planning than conversational `/plan`, especially when `prp-implement` will execute the result later.

0 installs
▌
Build Fix · elijj bundle
Use this skill when the user invokes `/build-fix`, says the build/typecheck/compile is failing, asks to fix TypeScript, Rust, Go, Java, Python, Gradle, Maven, package, import, or CI build errors, or when an implementation hits build failures. It fixes errors incrementally with minimal diffs, reruns the build after each fix, and stops when a fix requires architecture changes or missing dependencies.

0 installs
▌
Prp Commit · elijj bundle
Use this skill when the user invokes `/prp-commit`, asks for a smart commit, wants natural-language file targeting, needs staged/all/specific changes committed, or asks to create a conventional commit from current git changes. It interprets file selection, stages safely, writes a concise conventional commit message, and reports the commit hash and files.

0 installs
▌
Code Review · elijj bundle
Use this skill when the user invokes `/code-review`, asks for a code review, wants local uncommitted changes reviewed, provides a GitHub PR number, PR URL, or branch name, asks whether code is safe to merge, or after implementation before commit. It reviews correctness, security, type safety, tests, performance, and maintainability, runs applicable validation, and reports findings by severity before summaries.

0 installs
▌
E2e Testing · elijj bundle
Use this skill when creating, updating, debugging, or reviewing Playwright end-to-end tests, browser user flows, Page Object Models, CI E2E setup, flaky test handling, screenshots, traces, videos, or when a workflow invokes `/e2e`. It covers stable selectors, artifact management, critical flow coverage, and test reports.

0 installs
▌
Goal Grader · elijj bundle
Use this skill when the user asks to grade, review, score, improve, rewrite, harden, sanity-check, or optimize a Codex `/goal` prompt or any long-running autonomous Codex objective. Trigger for phrases like "grade this /goal", "is this a good goal", "optimize my goal", "score my autonomous task", "goal grader", or when a prompt starts with `/goal` and the user wants feedback before running it. The skill grades execution readiness, identifies blockers, and returns a safer optimized `/goal` rewrite.

0 installs
▌
Update Docs · elijj bundle
Use this skill when the user invokes `/update-docs`, asks to sync documentation from source-of-truth files, refresh generated docs, update command/env/API/contributing/runbook docs from code, or check docs for staleness after code changes. Converted from ECC `commands/update-docs.md` for Codex. The skill derives docs from scripts, schemas, routes, exports, env templates, and deployment files while preserving manual prose.

0 installs
▌
Tdd Workflow · elijj bundle
Use this skill when writing a new feature, fixing a bug, refactoring production code, adding an API endpoint, creating a component, or when the user invokes `/tdd` or asks for test-driven development. Trigger it even when the user only asks to refactor, fix, or implement something and the safest path is test-driven. It enforces RED to GREEN to REFACTOR with explicit test evidence, 80%+ coverage expectations, and unit/integration/E2E coverage scaled to risk.

0 installs
▌
Prp Implement · elijj bundle
Use this skill when the user invokes `/prp-implement`, provides a `.codex/prps/plans/*.plan.md` or other implementation plan file, asks to execute a plan step-by-step, or wants rigorous validation loops while implementing. It loads the plan, prepares git state, executes each task incrementally, runs validation after each meaningful change, writes an implementation report, and recommends `code-review` or `prp-pr` next.

0 installs
▌
Skill Creator · elijj bundle
Use this skill whenever the user wants to build, convert, test, benchmark, package, or improve a Codex skill. It applies to SKILL.md structure, frontmatter, trigger wording, bundled scripts/references/assets, agents/openai.yaml metadata, .skill packaging, trigger evals, baseline comparisons, review viewers, forward-testing, and iterative description optimization. Skip it for adjacent tasks like Python/npm packaging, GitHub Actions, generic model evals, plugin creation, README writing, or standalone scripts unless they are explicitly part of a Codex skill.

0 installs
▌
Update Codemaps · elijj bundle
Use this skill when the user invokes `/update-codemaps`, asks to update codemaps, generate architecture maps, refresh AI-readable architecture docs, scan project structure into `docs/CODEMAPS/`, or produce token-lean codebase maps after major feature work or refactors. Converted from ECC `commands/update-codemaps.md` for Codex. The skill scans project structure, writes compact codemap documents, detects large diffs before overwriting, and records a codemap update report.

0 installs
▌
Refactor Cleaner · elijj bundle
Use this skill when the user invokes `/refactor-clean` or asks to find and remove dead code, unused imports, orphaned files, duplicate logic, dead branches, or unused dependencies. It keeps cleanup cautious: classify risk, delete one item at a time, verify after each change, and stop when uncertainty appears.

0 installs
▌
Warp Worktree Fix · elijj bundle
Use this skill to repair a Warp terminal worktree launch config when the generated config itself is broken, misparsed, or incompatible with the target shell or OS. Choose it for Warp-authored worktree setup files that need path normalization, launch-entry cleanup, or safe preservation of placeholders and unrelated settings. Do not use it for ordinary git worktree commands, branch management, or general repo setup unless the Warp config is the thing that is failing.

0 installs
▌
Karpathy Guidelines · elijj bundle
Use this skill whenever writing, editing, reviewing, debugging, or refactoring code and the work could be over-scoped, ambiguous, or risky. This is the default coding-behavior layer for non-trivial work: make surgical changes, surface assumptions, avoid speculative abstractions, preserve existing style, and define verifiable success criteria. Trigger it for feature work, bug fixes, refactors, reviews, and any prompt asking for 'simple', 'minimal', 'don't overdo it', 'surgical', or 'Karpathy guidelines'.

0 installs
▌
Harness Skill Porting · elijj bundle
Use this skill only when the user wants to convert, migrate, port, translate, adapt, or get parity for a skill, prompt, command, agent, hook, rule, MCP config, or workflow from another agent harness into a Codex skill. This applies to Claude Code skills and slash commands, Cursor/Cline/Roo/Aider rules, prompt packs, and AGENTS/CLAUDE instructions. Trigger for phrases like "make this work in Codex", "full parity", "convert this skill", or "bring over this Claude workflow". Do not use for native Codex planning, PRP, TDD, build-fix, code review, PR, commit, E2E, packaging, or docs tasks unless the user is porting them from another harness.

0 installs
▌
Implementation Planning · elijj bundle
Use this skill whenever the user asks to plan before coding, uses `/plan`, requests an implementation plan, wants requirements restated, needs risks assessed, has ambiguous requirements, or is starting a complex feature, architecture change, multi-file refactor, migration, or high-risk change. Trigger it even when the user only describes the work and does not explicitly say "plan". This skill must produce a concrete implementation plan and stop for explicit confirmation before modifying code.

0 installs

17 published skills