superpowers
claudecodexCore skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Skill that enforces a doc-first per-phase workflow: brainstorm with research gate, planned execution with per-task cadence, dual-track smoke (manual + Playwright), phase delivery handoff, and milestone-done gate. Designed to be the same workflow across all your projects.
Plugin project-lifecycle from marketplace project-life-cycle in Siltpoke/project-life-cycle. Commands read from the repo's manifest — 19 days ago.
project-lifecycleUse when kicking off a new project, planning a new milestone, or executing per-phase work that needs traceable spec→plan→execute→journal→milestone-done discipline. Triggers include fresh project setup, "new milestone" requests, multi-task features, and decisions that warrant research before choosing
/project-lifecycle:builder-profileRead your own local Claude Code transcripts and produce a markdown snapshot of how you actually use an AI coding agent. 100% local — nothing leaves the machine. Gated pipeline (deterministic stats → evidence gate → cold-read → adversarial verify → independent verification). Default descriptive (mirr
/project-lifecycle:captureCapture a stated project intent (the "why") into the cold cognition log. Elicits at most two questions, then appends a validated entry to docs/cognition-log.d/<date>-<branch>.jsonl. Surfaces any prior intent on the same subject for supersede confirmation.
/project-lifecycle:catchupRender a warm "welcome back" card on return to a repo (post-/clear) — where you are, what shipped, what you were doing, and whether your save is fresh — synthesized from live git + RESUME.md + PLC's session digest (+ STATUS.md if present). AI-rendered normal reply, read-only, omits rows it has no da
/project-lifecycle:cognition-distillRegenerate the hot cognition doc (docs/cognition.md) from the cold intent-log — reads the log, not the prior doc (regenerate-from-source). Groups non-superseded intents into four sections, one fact + source pointer per line, then runs the deterministic guards (freshness, cap, dead-path, coverage, sn
/project-lifecycle:experimentWrite an experiment proposal for this phase, one section at a time, stopping for the user at each. Nine sections — Hypothesis / Baseline / Dataset / Ground Truth / Metrics / Procedure / Variance / Decision Rule / Cost — landed as a file under docs/experiments/. Per-phase step 3.5. It does NOT run th
/project-lifecycle:handoffWrite a schema-enforced, MOMENT-tense mid-session continuity snapshot to RESUME.md (+ a conditional journal FACT) by following references/handoff-snapshot.md. Captures the invisible state git can't recover — current task+phase, next concrete action, staleness anchor, and (when non-empty) decisions /
/project-lifecycle:init-harnessBootstrap a project to use the project-lifecycle skill — detect stack + folder layout + tooling, generate CLAUDE.md / folder-map / policy keys / project-shared commands / deterministic handler scaffolds. Idempotent (merges into existing files; never overwrites without explicit confirm). 4 human chec
/project-lifecycle:plc-statusRender the lifecycle ledger — for every mandated stage of a run, whether it was observed, only claimed, missing, not-applicable, or unrecoverable. Default shows the current run, `--run <id>` shows one run's rows plus its timeline, `--all` shows the conformance matrix over every run on disk. Read-onl
/project-lifecycle:recallBrief the current session from the most-recent PLC session digest for THIS repo (written automatically by the save_session Stop hook to ~/.claude/plc-session-data/). Read-only — surfaces where you left off (tasks in flight, files touched, tools used) and stops; never auto-starts work or edits files.
/project-lifecycle:reconcileReconcile conversation-made roadmap/status decisions (and merged PRs since the last checkpoint) against docs/ROADMAP.md + the project's read-first status doc (`RESUME.md` — `docs/STATUS.md` only where an adopter has pointed `status_doc` at one; see `references/document-map.md`). Proposes staged, ind
/project-lifecycle:releaseCut a new release end-to-end. Computes SemVer bump from CHANGELOG [Unreleased], renames the section, bumps all plugin manifests, validates, commits, tags, pushes, and verifies the GitHub Release landed. One human checkpoint (confirm bump + version).
/project-lifecycle:researchRun PLC's brainstorm research protocol on a single question OUTSIDE a brainstorm, and write a cited report to docs/research/. Dispatches parallel Tier-1/Tier-2 survey agents, synthesizes a cited recommendation, and (by default) runs a blind 2nd-agent verification. Produces a research artifact — it d
/project-lifecycle:reviewRun a PLC-native code review on the current branch's diff (or a given path/range) by dispatching a general-purpose agent briefed with references/reviewer-brief.md. Produces findings under the brief's format — file:line evidence, refute-first, verdict-LAST. It does NOT compute its own approve/block v
/project-lifecycle:shipShip one vertical-slice feature end-to-end via the project-lifecycle factory chain. Chains researcher → story → spec → BE builder → FE builder → acceptance verifier → validator → fix loop → PR. 3 human checkpoints (story / spec / PR).
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development
9 skills · 1 command · 8 agents
AI-supervised issue tracker for coding workflows
1 skill · 1 agent · hooks
MiniMax AI skills library for frontend, fullstack, Android, iOS, shader, GIF sticker, document, presentation, spreadsheet, and multimodal media workflows
18 skills
Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows. This plugin provides configuration guidance (LLM commit messages, project hooks, worktree paths), automatic activity tracking (🤖/💬 indicators in `wt list` showing active Claude sessions), and a `/wt-diff` pane showing each worktree's changes.
2 skills · hooks
GSD Core is a meta-prompting, context engineering, and spec-driven development system for AI coding agents.
72 skills · 72 commands · 64 agents · hooks
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.