Spec-Driven Development for Claude Code: 22 atomic, stack-agnostic skills + 11 subagents that carry a feature from idea → spec → ux-flows → design → data-model → API → screens → tasks → tests → implement → review → ship, plus a TDD implementation engine and a living decomposition roadmap above the per-feature flow.…
Chart a big, foggy effort as a shared map of investigation tickets under .plan/, resolved one per session. Bundles the wayfinder method plus the four skills it invokes: grill-me, research, prototype, and domain-modeling. Skills originally authored by Matt Pocock (github.com/mattpocock/skills), adapted here.
PRD operating system: capture rough ideas, draft reviewable PRDs, run standard and adversarial review (Codex or a Claude senior-staff-engineer subagent, recorded truthfully), triage findings, decompose approved PRDs into atomic issue specs, and execute those issues with scope enforcement, stop-gate receipts, and…
108
changed yesterdayA
tokens not measured
originalMIT
Universal radial-tree exploration engine for Claude Code. One tree skill + swappable presets (brainstorm / attack / design / code-audit) for divergent ideation, adversarial critique, and design-space exploration. 12 framings × hard-ban-on-incomplete-leaves × stable convergence. See CHANGELOG.md for per-version histor.
MBB-level management consulting skill for Claude Code. Strategic frameworks (MECE, Porter's Five Forces, McKinsey 7S, BCG Matrix), hypothesis-driven problem solving, case interview structures, consulting deliverable templates, and real-world case studies across industries.
Architecture Deep Research — scan a repo, draft a PRD, run live deep research, get a cited handoff. Use when picking a topology, retrieval architecture, event bus, storage engine, auth pattern, or any system-design decision before code lands.
105 2mo agoA
tokens not measured
originalApache-2.0
Draft or activate compact goals for verifiable long runs. Defines the result, evidence, boundaries, stop rules, and only useful delegation; separate tasks remain user-directed.
Self-enforcing engineering methodology extracted from Claude Fable 5 for less advanced models: prime directives + integrity rules injected each session, 26 on-demand skills, 4 contracted subagents (builder, qa-verifier, code-reviewer, research-scout), and 6 lifecycle hooks that gate destructive commands and unverified.
Loops, swarms, and teams powered by Claude Code's built-in Task System. Plans define exit criteria. Executors run until tests pass. Done means actually done.
A suite of composable skills (Claude + OpenAI Codex) for taking an idea from fuzzy to validated to sequenced build to shipped, by hand or on autopilot: a manual tier (ideate, deep-dive, prompt-pack) and an autonomous tier (autopilot, build-loop, audit-and-fix).
83 29d agoA
tokens not measured
originalMIT
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: