Project-enabled scaffolding layer: the milestone substrate (playbooks, forward-looking capture, the post-merge status sweep) and the second-recurrence skill-authoring trigger, the ADR + milestone commands and the /scaffold:milestone-run runner, context-checkpoint, config-guard, and trunk-write-guard (block-main-writes) hooks, a determinism auditor, and references/ for by-hand project setup. Complements the superpowers plugin (which owns plan execution and skill authoring) rather than duplicating it. Enable per-repo in .claude/settings.json; requires the core plugin and assumes the docs/ + .context/ layout.
Scaffold the next ADR in docs/decisions/ from TEMPLATE.md
Dispatch a milestone build end-to-end (drives /milestone via the milestone-workflow skill)
Print the exact terminal command to run the milestone loop (resolves the scaffold plugin path).
Generic per-milestone driver (copy into named /m1, /m2, … as milestones firm up)
Use at the start of and throughout any milestone build — the project-substrate half of a milestone: playbook + preconditions, forward-looking data capture, the in-PR context rule, and the status-prose sweep at milestone completion. Trigger from the /goal and /milestone commands. Defers plan execution to superpowers:executing-plans (or subagent-driven-development) and the evidence-before-done gate to superpowers:verification-before-completion.
Use to decide WHEN a project skill is worth authoring — the second-recurrence trigger — and what belongs in one. Trigger when the same knowledge need (a data shape, an API, a procedure) comes up a second time. Covers the trigger only; superpowers:writing-skills owns how to actually write and test the skill.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A personal Claude Code plugin marketplace — portable personal skills, agents, hooks, and an output style, version-controlled, with zero manual sync across machines and repos. The idiomatic vehicle for keeping a personal steering layer in sync (per code.claude.com/docs: plugins, plugin-marketplaces, skills).
| Plugin | Scope | Contents |
|---|---|---|
core | user-level (all repos) | 4 skills (cost-aware-delegation, orchestration, git-workflow, deep-research-tiered), 5 agents (advisor-plus, code-reviewer, search, doc-sync, config-auditor), the distinguished-engineer output style, 1 guard hook (guard-secrets) |
scaffold | per-repo (opt-in) | 2 skills (milestone-workflow, skill-maintenance), 4 commands (/adr, /goal, /milestone, /scaffold:milestone-run), the determinism-auditor agent, 4 hooks (block-main-writes, checkpoint, subagent-trail, validate-config), and references/ for by-hand project setup. Requires core. |
core is broadly/globally enabled and project-agnostic. scaffold is the project-enabled half: turn it on
per-repo, and it assumes a concrete docs/ + .context/ layout (hand-set up per
scaffold/references/project-setup/ — there is no generator).
scaffold declares a hard dependency on core in its plugin.json ("dependencies": ["core"]), so
enabling scaffold auto-enables core. Each plugin's own README is the authoritative inventory. Further
plugins are added as new entries in .claude-plugin/marketplace.json.
TODO —
/scaffold:adoptgenerator: replace the by-handscaffold/references/project-setup/steps with a single command that bootstraps a repo's.context/, ADR conventions, andsettings.jsonfrom templates bundled in the plugin — one command instead of hand-copying, updatable via/plugin update. Not built yet; the by-hand setup works today.
.claude/skills/), and Apple's bundled Xcode skills (swiftui-specialist, etc.) that are re-exported per
Xcode update and would go stale if committed — keep those at user level.superpowers already does better — see below.superpowerssuperpowers@claude-plugins-official is assumed present and is
the process layer: brainstorming → writing-plans → executing-plans / subagent-driven-development →
TDD → systematic-debugging → verification-before-completion → writing-skills. This marketplace does
not re-implement any of it. The division:
| Concern | Owner |
|---|---|
| How to write, test, and structure a skill | superpowers:writing-skills (+ skill-creator) |
| Executing a written plan task-by-task; dispatching per-task subagents | superpowers:executing-plans / subagent-driven-development |
| How to word a subagent dispatch (pack context, scope, output format) | superpowers:dispatching-parallel-agents |
| Evidence before claiming done | superpowers:verification-before-completion |
Which model runs it, and what it costs — concrete Haiku/Sonnet/strong routing, opts.model per Workflow stage, the budget guard, checkpoint-before-loss | core — see the note below; superpowers tiers by role but never names models or Workflow primitives |
| Trunk-based git policy — the PR merge method (rebase vs squash), branch routing, protection, required checks, CI delegation | core:git-workflow — superpowers defines only a local git merge in finishing-a-development-branch, and never a GitHub PR merge method |
Milestone substrate (docs/playbooks/, .context/, the post-merge status sweep) | scaffold:milestone-workflow |
| Agents, commands, output style, guard hooks | core/scaffold — superpowers ships none of these |
Cross-repo personal Claude Code layer: skills (cost-aware delegation with a Sonnet subagent floor, tiered orchestrator-worker fan-out, trunk-based git workflow, tiered deep research), review/search/doc agents, the distinguished-engineer output style, and a universal secret-guard hook (the trunk-write guard moved to the project-scoped scaffold plugin). Owns model tiering and git-workflow policy; defers prompt craft, plan execution and skill authoring to the superpowers plugin.
npx claudepluginhub mrinaldhillon/claude-skills --plugin scaffoldHarness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
v9.54.1 — Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques