Loading the repository README…
Showing 40 of 98 collected skills.
Land the current work as atomic Conventional Commits with a DCO Signed-off-by trailer — scope intake, stage-split when concerns mix, message draft, sign and write. Use when a verified change is ready to land or on "commit this", "stage and commit", "split…
Land the current work as atomic Conventional Commits with a DCO Signed-off-by trailer — scope intake, stage-split when concerns mix, message draft, sign and write. Use when a verified change is ready to land or on "commit this", "stage and commit", "split…
Scaffold a staged spike with golden fixtures when the spec is clear but the technique is uncertain — discovery, fixture, gated pipeline with per-stage debug artifacts, two-layer evaluation. Use on "spike", "which library", "evaluate approaches", "golden…
Generate or audit ARCHITECTURE.md at the repo root: scans the code, pre-fills layers, patterns, observability, and deployment from observed signals, then asks only the real gaps. Use to bootstrap, scaffold, document, or audit ARCHITECTURE.md (system patterns…
Hard-delete completed plan files (done tasks, shipped specs, superseded PRDs, superseded or deprecated ADRs) with `git rm`, leaving git history as the ledger. Use to archive, clean, prune, or sweep finished records out of the tree. An accepted ADR goes only…
Generate or audit AGENTS.md at the repo root: scans the codebase, pre-fills from observed signals, and asks only the real gaps. Use to bootstrap, scaffold, create, set up, or audit AGENTS.md or CLAUDE.md, the operational guide for agents on the project.…
Scaffold or audit GitHub-conventional CONTRIBUTING.md and SECURITY.md without inventing maintainer policy or overwriting existing documents. Use when a maintainer needs contributor guidance, a security-reporting policy, community health files, contribution…
Generate or audit DESIGN.md from existing design tokens (Figma export, tailwind.config, tokens.json, stylesheets) as W3C-compatible YAML plus Markdown rationale, per the Google Labs design.md spec. Use to bootstrap, document, or audit the visual contract:…
Draft, update, or audit GUIDELINES.md, the full engineering reference: design principles, code standards, complexity, API, build, static analysis, quality gates, testing, git, documentation, security. Scans the toolchain and config first, then asks only real…
Scaffold deterministic quality gates: pre-commit (lint, format, secret-scan) and pre-push (build, tests) with a hook runner fitted to the stack, plus a PostToolUse artifact-validator gate on both hosts and the Stop nudge and workflow checkpoint on Claude…
Curate the project's rule-set (add, refine, merge, or retire a convention) through four anti-overfitting gates, an effectiveness pass, and an adversarial multi-lens review; writes one item at a time, only on the owner's explicit approval. Use on "add a…
Draft or review collaboration publications (a GitHub proposal issue, a Slack, Discord, issue, or pull-request reply) from approved templates, grounding claims and keeping private context out of outward text. Posts only after an exact-target approval from the…
Orchestrate one npm package release (release script, release PR, tag, npm publish, GitHub Release) behind one digest-bound approval from the owner, with per-stage fallback and state-aware resume. Use on "release this package", "publish to npm", "prepare a…
Draft, review, render, or verify a standalone analytical report from approved templates (Markdown, HTML, DOCX, PDF, slides, visualizations), preserving facts and asking before any export beyond the agreed local target. Use when the user wants a report from…
Draft a new Claude Code or Codex skill at .claude/skills/<name>/SKILL.md or .agents/skills/<name>/SKILL.md in the Anthropic Skills format. Use when the user wants to create, write, draft, or scaffold a custom skill for an agentic coding tool. Asks one…
Draft a Codex custom subagent at .codex/agents/<name>.toml in the official format, for delegated work such as a fresh-context reviewer, codebase explorer, docs researcher, test designer, bug reproducer, or bounded worker. Use to create, write, or scaffold a…
Curate durable publication and report templates from authorized examples, keeping owner, community, external, and official provenance apart and storing derived rules, not raw sources. Proposes one delta at a time and writes only the candidate the owner…
Preview and safely update the machine-global Agentic Development kit. Use when the user asks to update agentic, refresh installed skills, get the latest kit, sync the engineering skills, inspect kit drift, or invokes `/ad-update`. Project scope is explicit;…
Draft, rewrite, or translate text through the owner's confirmed machine-local voice profile, preserving facts and applying attributed team or community accommodation. Read-only: never changes the profile. Use on "write like me", "use my voice", "make this…
Create or refine the owner's machine-local voice profile from owner samples or authorized conversations, separating identity from community accommodation and storing derived patterns. Negotiates sources before private reads, proposes one delta at a time,…
Generate or audit ARCHITECTURE.md at the repo root: scans the code, pre-fills layers, patterns, observability, and deployment from observed signals, then asks only the real gaps. Use to bootstrap, scaffold, document, or audit ARCHITECTURE.md (system patterns…
Hard-delete completed plan files (done tasks, shipped specs, superseded PRDs, superseded or deprecated ADRs) with `git rm`, leaving git history as the ledger. Use to archive, clean, prune, or sweep finished records out of the tree. An accepted ADR goes only…
Generate or audit AGENTS.md at the repo root: scans the codebase, pre-fills from observed signals, and asks only the real gaps. Use to bootstrap, scaffold, create, set up, or audit AGENTS.md or CLAUDE.md, the operational guide for agents on the project.…
Scaffold or audit GitHub-conventional CONTRIBUTING.md and SECURITY.md without inventing maintainer policy or overwriting existing documents. Use when a maintainer needs contributor guidance, a security-reporting policy, community health files, contribution…
Generate or audit DESIGN.md from existing design tokens (Figma export, tailwind.config, tokens.json, stylesheets) as W3C-compatible YAML plus Markdown rationale, per the Google Labs design.md spec. Use to bootstrap, document, or audit the visual contract:…
Draft, update, or audit GUIDELINES.md, the full engineering reference: design principles, code standards, complexity, API, build, static analysis, quality gates, testing, git, documentation, security. Scans the toolchain and config first, then asks only real…
Scaffold deterministic quality gates: pre-commit (lint, format, secret-scan) and pre-push (build, tests) with a hook runner fitted to the stack, plus session hooks: a Stop handoff nudge, a UserPromptSubmit workflow checkpoint, and a PostToolUse…
Curate the project's rule-set (add, refine, merge, or retire a convention) through four anti-overfitting gates, an effectiveness pass, and an adversarial multi-lens review; writes one item at a time, only on the owner's explicit approval. Use on "add a…
Draft or review collaboration publications (a GitHub proposal issue, a Slack, Discord, issue, or pull-request reply) from approved templates, grounding claims and keeping private context out of outward text. Posts only after an exact-target approval from the…
Orchestrate one npm package release (release script, release PR, tag, npm publish, GitHub Release) behind one digest-bound approval from the owner, with per-stage fallback and state-aware resume. Use on "release this package", "publish to npm", "prepare a…
Draft, review, render, or verify a standalone analytical report from approved templates (Markdown, HTML, DOCX, PDF, slides, visualizations), preserving facts and asking before any export beyond the agreed local target. Use when the user wants a report from…
Draft a new Claude Code or Codex skill at .claude/skills/<name>/SKILL.md or .agents/skills/<name>/SKILL.md in the Anthropic Skills format. Use when the user wants to create, write, draft, or scaffold a custom skill for an agentic coding tool. Asks one…
Draft a Claude Code subagent at .claude/agents/<name>.md in the official format, for delegated work such as a fresh-context reviewer, codebase or docs researcher, test designer, bug reproducer, or bounded worker. Use to create, write, or scaffold a custom…
Curate durable publication and report templates from authorized examples, keeping owner, community, external, and official provenance apart and storing derived rules, not raw sources. Proposes one delta at a time and writes only the candidate the owner…
Preview and safely update the machine-global Agentic Development kit. Use when the user asks to update agentic, refresh installed skills, get the latest kit, sync the engineering skills, inspect kit drift, or invokes `/ad-update`. Project scope is explicit;…
Draft, rewrite, or translate text through the owner's confirmed machine-local voice profile, preserving facts and applying attributed team or community accommodation. Read-only: never changes the profile. Use on "write like me", "use my voice", "make this…
Create or refine the owner's machine-local voice profile from owner samples or authorized conversations, separating identity from community accommodation and storing derived patterns. Negotiates sources before private reads, proposes one delta at a time,…
Curate the project's rule-set (add, refine, merge, or retire a convention) through four anti-overfitting gates, an effectiveness pass, and an adversarial multi-lens review; writes one item at a time, only on the owner's explicit approval. Use on "add a…
Curate the project's rule-set (add, refine, merge, or retire a convention) through four anti-overfitting gates, an effectiveness pass, and an adversarial multi-lens review; writes one item at a time, only on the owner's explicit approval. Use on "add a…
Maximum-gate adversarial audit of a diff, branch, PR, or drafted claims against the repo's rule-set — on Codex one inline pass and audit trail, every rule given a verdict. Use before posting to the team, opening a PR, or handing off; "audit against the…
Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/alexandremendoncaalvaro/agentic-development.Repository README source: README.md
Read the source README