Skill Codex
Use when the user asks to plan a feature before coding, break an epic or outcome into implementation-ready work, define behaviour and acceptance criteria, write a technical plan, or create issue-ready dependency slices. Produces feature contracts, implementation plans, work packages, and acceptance artefacts that can…
Skill Claude CodeCodex
Use when the user asks to start, continue, coordinate, or close a repository engineering session and the work may span planning, tasks, implementation, worktrees, documentation, change explanation, or handoff. Identifies the current workflow stage, routes to the right specialist skills, and keeps session state and…
Skill Codex
Use when the user is stopping work for now, asks to leave a handoff for the next session, or needs current progress, decisions, risks, and next actions preserved locally. Creates one active dated continuation handoff outside canonical knowledge and task bundles, then manages supersession and cleanup. Runs after RSA…
Skill Codex
Use when the user asks to resume, pick up, or continue work from a local handoff, or a new session begins with an existing handoff file. Selects the one active handoff, rebuilds trustworthy repository context before editing, ignores superseded records, and re-verifies stale claims. Do not use to create a new handoff…
Skill Codex
Use when material implementation has finished or the user asks what changed, how a new path works, what calls what, what was removed, why an implementation changed, or what they need to understand before committing. Verifies the bounded diff and explains the causal symbol-level path for commit context and the closing…
Skill Codex
Use when the user inherits an unfamiliar, undocumented, unclear, or vibe-coded repository; asks how the whole repo works; wants entry points, runtime paths, state, boundaries, and drift mapped; or needs the first trustworthy documentation foundation. Verifies code, runtime, and existing docs rather than trusting…
Skill Codex
Use when the user says repository docs are stale, wrong, duplicated, scattered, or hard to search; asks where something is documented; wants a decision or glossary term written down properly; or when RCC exposes a gap in canonical docs. Maintains the trusted documentation, reading order, decisions, glossary, and…
Skill Codex
Use when the user asks to finish, finalise, harden, tidy, publish, open-source, or make a software repository public-ready. Closes development loose ends, aligns canonical docs, removes stale surfaces, checks active paths and TODOs, and runs bounded secret, PII, credential, local-path, environment-file, cache, and…
Skill Codex
Use when about to report completion, hand off, or close a material engineering session and repository tasks, canonical docs, evidence, or the user-facing explanation may have drifted from what actually landed. Reconciles task execution truth, RKE-managed knowledge, validation evidence, and the verified RCC summary…
Skill Codex
Use when the user asks to set up, initialise, or scaffold a new repository; choose a licence; add CODEOWNERS, contributor or agent instructions, branch protections, repository rulesets, release scaffolding, or a GitHub description; or establish baseline governance before engineering begins. Produces the governed…
Skill Codex
Use when the user asks to create or update repository-local tasks, workstreams, status, estimates, time, evidence, backlog, or execution history that must survive chat; wants a task board or graph; or needs safe tracker import or sync. Maintains the durable task ledger and, where adopted, its OKF Tasks bundle and…
Skill Codex
Use when the user asks for a QA test plan, test cases, requirement-to-test coverage, acceptance-test scenarios, or a review of gaps in an existing test set based on requirements, stories, change notes, or specifications. Produces proportionate human-reviewable test coverage and traceability. Do not use for coding…
Skill Codex
Use when the user asks to publish, create, or synchronise already-defined work packages or repository task records into GitHub, GitLab, Linear, ClickUp, or another external tracker. Adapts the existing hierarchy and acceptance surface to the target tracker using shared output defaults. Do not use to design the work…
Skill Codex
Use when the user asks to split work across Git worktrees or branches, coordinate parallel or stacked changes, assign path ownership, track branch dependencies, reconcile confusing managed worktrees, or clean them up after merge. Maintains explicit worktree, branch, dependency, integration, and cleanup records. Do not…
Skill Codex
Use when the user asks to make narration TTS-ready for ElevenLabs, clean a voiceover script, improve spoken pacing, handle pronunciation or acronyms, or add selective ElevenLabs v3 voice tags. Produces a cleaned, speakable script without changing the intended message. Do not use for generating audio itself or for full…
Skill Codex
Use when the user asks to build an illustrated, walkthrough, or hybrid explainer video in Remotion; needs scene prompts, composition plans, timings, overlays, branded shells, debug compositions, or guidance combining stills and screen recordings. Produces the multi-pass video design and Remotion implementation plan.…
Skill Codex
Use when the user asks to fix, repair, rewrite, or harden a prompt, skill, system instruction, agent prompt, or related LLM instruction artefact and already has findings, an issue list, or a clear failure brief. Applies conservative reliability fixes without changing the artefact's purpose. Do not use when the user…
Skill Codex
Use when the user asks to review, critique, audit, or pressure-test a prompt, skill, agent instruction, or system prompt before publication or repair; or asks why an instruction artefact behaves inconsistently. Reports execution risks and repair directions without rewriting the artefact. Do not use when direct fixes…
Skill Codex
Use when the user asks to set up or refresh shared Polaralias defaults; configure branding, fonts, assets, tone, continuity, tracker, or structured-output preferences across repositories; or install portable Claude or agent routing instructions for the Polaralias skill families. Writes update-safe user or project…
Skill Codex
Use when the user asks to finalise, normalise, package, publish, or import a skill; extract a zipped skill; fix SKILL.md frontmatter or agents/openai.yaml; add its licence, icon, tests, or references; or validate package hygiene before release. Produces a clean publishable skill package. Do not use to design a new…
Skill Codex
Use when the user asks an unresolved repository question, wants a term or decision clarified, says the docs and code disagree, asks which source is right, or needs assumptions pressure-tested before they become durable truth. Investigates the largest useful related question set and records resolved answers in the…
Skill Codex
Use when the user asks to understand or explain repository code or a diff, implement or refactor changes, perform code review or testing, review documentation, plan acceptance, reconcile tasks and knowledge, resume or coordinate work, assess release readiness, or prepare publication. Selects proportionate engineering…
Skill Codex
Use when the user asks to test whether a skill or LLM instruction triggers correctly, avoid false positives, build a routing or behaviour scenario matrix, choose graders, create an evaluation plan, or emit runner-specific output such as Waza. Produces activation, non-activation, success, boundary, and regression…
Skill Codex
Use when the user asks to install, configure, qualify, troubleshoot, or use the RKE repository runtime, or when an engineering method needs available RKE retrieval, structure, knowledge-validation, or continuity support. General engineering reasoning belongs to Engineering Workflow and does not require RKE. Runtime…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: