adr plugins

by Rob Sartin

Scaffold a stack-appropriate docs/adr/ from composable ADR packs.

by taika-izumi

AI駆動開発ガイドラインを実装するスキル群（start-work, decision-log, session-handoff, feature-block-design, retrospective ほか）。導入先プロジェクトの AGENTS.md と組み合わせて使用する。

by taika-izumi

AI駆動開発ガイドラインを実装するスキル群（start-work, decision-log, session-handoff, feature-block-design, retrospective ほか）。導入先プロジェクトの AGENTS.md と組み合わせて使用する。

by Chadwick Puterbaugh

Principal-level architecture consultant subagent team that runs a gated discovery-to-brief workflow — with a skeptic/practical-planner check-and-balance and a cartographer that diagrams every architectural output — without touching implementation.

by dimonb

Run a multi-agent discussion room: several agent sessions (Claude Code, Codex, Antigravity) argue one question in background terminals by a strict protocol — speaking in turn, objecting with references, closing objections mechanically — and the room writes a decision record when it converges, or an honest unresolved one when it does not.

by HernandoR

ADR-driven development, agent spec conventions, decision grilling, English-only artifacts, controlled-language rewrites, and diff explanations

by claude-skills

Keeps repository documentation consistent and honest. Scaffolds a universal document set scaled by repo archetype, audits existing repos for missing, stale, or off-standard docs, enforces append-only architecture decision records, compacts the oldest of them into a digest once fifty exist, and regenerates the generated ones so drift from the code is reported rather than discovered. Ships a plain-English Vale style alongside it, plus optional compliance overlays for IEC 62304, the OSPS Baseline, the EU Cyber Resilience Act and NIST SSDF, each describing document structure only and never inventing regulatory content.

by djjay0131

Executive AI personas (incl. Repository Steward) and the governance establish/audit/migrate skills for agentic-governance adopters

by Carlos Araujo

Entrega hotfixes por uma faixa rápida e verificável; reconciliação incremental com receipts fail-closed.

by KazzerLabs

Memory and governance for Claude Code: a read-before-write gate, a context resolver with receipts, tiered project notes, handoffs between sessions, and docs checks that fail when the record and the tree disagree.

by kgai maintainers

Shared decision memory for AI dev teams. Your agent records the decisions behind your code — what changed, why, even the dead ends you ruled out — recalls them before changing code, and syncs them across the team without merge conflicts. Local-first, immutable, measured to 1M decisions.

by Arbel Hakopian

Checked memory skills for AI coding agents. Resume anywhere: any task, any session, any model, from a checkout. One markdown record per task, committed like code and checked like code, with a lint, a review gate, decision records, a generated HANDOFF page, and session archives.

by Calvin Dittkrist

The repository standard: AGENTS.md with a CLAUDE.md import, the make check gate, architecture.md, ADRs, PR template, workflow settings, and a check that fails on stray agent configuration. Skills: /standardize (six read-only auditors, one findings report), /apply (backup, cleanup pull request, issues, GitHub workspace), /adr, /docs-check.

by Maryan08

AI Agentic Workflow for Claude Code: nineteen atomic, stack-agnostic skills + subagents that carry a feature from idea → spec → design → data-model → API → tasks → tests → implement → review → ship, plus a TDD implementation engine and a living roadmap. AI is the first lens (design through it, not bolt it on); the engineer stays architect and quality controller (every stage is Socratic and gated, not a wall of output); a domain anchor grounds the work (survey maps the repo before anything else); verification is a loop, not an afterthought (plan-tests maps every acceptance criterion to a test, implement is TDD, review runs in clean context so whoever wrote the code doesn't approve it alone); and honesty about uncertainty is built in (clarify runs a devil's-advocate pass before code exists). Start with survey, walk the gated backbone — each stage reads the previous artifact and hard-refuses if missing — keep the Now/Next/Later roadmap in sync, and codify what you learn back into the shared library.