Agents-spec

Discovers and invokes agent skills. Use when starting a session or when you need to discover which skill applies to the current task. This is the meta-skill that governs how all other skills are discovered and invoked.

v0.1.0 - Delivers changes incrementally. Use when implementing any feature or change that touches more than one file. Use when you're about to write a large amount of code at once, or when a task feels too big to land in one step.

v0.1.0 - Breaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible.

v0.1.0 - Simplifies code for clarity. Use when refactoring code for clarity without changing behavior. Use when code works but is harder to read, maintain, or extend than it should be. Use when reviewing code that has accumulated unnecessary complexity.

v0.2.0 - Draft, split, and execute scoped Git commits with issue traceability and structured Why/What/Impact/Tests/Refs bodies; use when preparing final commits, reviewing commit wording, or enforcing repo commit conventions.

v0.1.18 - Check, create, and link GitHub or GitLab issues for tracked work before implementation and before commit preparation; use when a repo requires issue-backed traceability, canonical upstream issue confirmation, product-facing parent issue framing, child issue drafting for engineering detail, fork-vs-upstream issue targeting, or commit Refs output.

v0.1.0 - Evaluate GitHub PRs or GitLab MRs, normalize review findings, and optionally publish review comments or verdicts via gh/glab.

v0.1.2 - Analyze one source branch diff, split independent change slices into separate branches, and open one GitHub PR or GitLab MR per slice when personal work was developed together on one branch.

v0.1.2 - Prepare, execute, and verify straightforward tag and hosted release operations from explicit operator inputs.

v0.1.2 - Produce a document-first, from-scratch implementation guide that derives one feature or method from behavior and invariants into step-by-step structure, helper contracts, small code fragments, and verification; use it when the human wants to learn how to build the core logic rather than trust final code.

v0.1.20 - Guide one human-led, main-first landing wave from an AI draft or worktree, keeping `Human-Owned` core logic on `main`, AI output as reference only, and verification plus `commit_when` checkpoints explicit.

v0.1.4 - Distill a feature or system core into a runnable minimal-complete reference sample before `main` integration; use it when production code or an AI draft is too noisy to learn from safely, then hand off to human-led main landing.

v0.1.8 - Create a project-specific Codex skill package with native Codex defaults and master-grade governance; use when building reusable skill folders, onboarding skills, coordination skills, or general project skills that should ship as native Codex skills (not API-only), including optional paired reviewer skills for high-freedom outputs.

v0.4.12 - Record one concrete documentation artifact at the correct code-owned node with reader-first page choices, stable topic-first filenames, topic-tree role priority, reader-facing section titles, single-job page boundaries, node-local topic containers, README-vs-docs separation, and preserved local detail so it fits the repository's book-like documentation system.