research
claudeMulti-phase research workflow for building cited, audited knowledge bases.
21 skills · 10 agents
Demo-driven, proof-driven build workflow: vision kickoff, demos as executable human checkpoints (a scenario plus an end-to-end check that alone turns them green), tasks in one flat dependency graph, decisions with explicit proof obligations, parallel TDD burn-down with grading and demo checks after merges, and a verified human test cycle.
Plugin trajectory from marketplace mtrense-skills in mtrense/skills. Commands read from the repo's manifest — today.
aimDefine the next demo of a trajectory project — the next point where something can really be shown working to a human: take an idea, bullet points, or an example (or propose the next demo yourself when invoked empty), scout the current project state via subagents, and write demos/NNNN-slug.md coverin
burnBurn a trajectory backlog down: drive available tasks (todo with all dependencies done, ascending id) through parallel burn-worker subagents in git worktrees — strict TDD, commit via /commit — then grader → serialized closing-record scribe → sequential merge → serialized doc-sync → demo checks for t
decideRecord, revise, or supersede one decision in a trajectory project's decision log. Takes a statement to decide on, a reference to an existing decision (to revise or supersede), or — when invoked empty — the next open decision item from the lowest-id open demo. Runs a brief Socratic dialog to a shared
enrichBreak a trajectory demo down into well-shaped tasks: a dialog that turns the next (or given) demo's scenario into tasks/NNNN-slug.md files — the demo's executable check first, then the work that makes it pass — each with a plan, decision/documentation references, complexity, dependencies, acceptance
kickoffStart a trajectory project (or revise its vision): a Socratic interview about users and their needs, what success looks like, hard constraints/invariants, and the domain vocabulary, wrapped up as VISION.md with an explicit ## Vocabulary section. Trigger on "/kickoff", "start a new project", "let's d
landRun the human test cycle on a green trajectory demo: re-run its check at HEAD, compose a verified walk-through (setup, numbered steps with expected observations, the human-only checks, deviations and caveats) from the demo's scenario and the closing records of the tasks behind it, let the user try i
supplementAdd a small demo-less task to a trajectory backlog — a bug, a small improvement, a chore — with the same rigor as /enrich: plan, decision/doc references, complexity, dependencies, acceptance criteria, duplicate/link scan, and vocabulary check, including proposing a breakdown into several tasks when
aim-scoutRead-only project reconnaissance for the trajectory /aim skill. Given a demo idea and the repo root, surveys the current state of the codebase and backlog relevant to that idea: what already exists, what the idea would touch, the real entry points a demo scenario could drive, unknowns that need a de
burn-scribeClosing-record scribe for the trajectory /burn skill. After a grader accepts a task and before its merge, reads the worker's full report from .burn/REPORT.md in the task's worktree and writes the closing record — the ## Manual testing and ## Deviations sections — into the mainline task file, committ
burn-workerSingle-task TDD implementation worker for the trajectory /burn skill. Receives pointers — a dedicated git worktree and the task file's path inside it — and self-briefs: reads the task body (plan, acceptance criteria, grader feedback if this is a retry) and only the linked decisions' digest lines its
decision-summarizerWrite-side worker that keeps a trajectory project's per-topic decision digests in sync with its decision records. Given the documentation directory and the id(s) of the decision(s) just recorded or revised, reads those records, maps each to one or more topics (tech-stack, persistence, testing, error
doc-syncerPost-merge documentation refresher for the trajectory /burn skill. After one task's merge lands, checks whether that change made any documentation stale — README usage, reference docs, examples, the task's listed documents — and makes the minimal updates. Runs strictly serialized (one at a time, in
example-drafterRead-only strawman drafter for the trajectory /example skill. Given an intent (what artifact, pinning what), the project root, and a mode — draft, annotate (an artifact the user provided), or revise (an existing example) — reads the vision vocabulary and invariants, the decision index and relevant t
graderStrict acceptance grader for the trajectory /burn skill. Given pointers — the task file's path, the worker's worktree + commit, and the worker's report file — self-briefs on the acceptance criteria and linked decision digests, then judges the output against those criteria and decisions — nothing els
mergerSequential merge-back worker for the trajectory /burn skill. Given one accepted task's worktree branch and the main workspace, merges the task's commit into the mainline, re-runs the test suite post-merge, and cleans up the worktree on success. A textual merge conflict is a bounce, not a quality pro
task-linkerRead-only duplicate-and-link scan for the trajectory /enrich and /supplement skills. Given one proposed task (title, plan summary, acceptance criteria) plus the current backlog, finds existing open tasks that already cover part of the proposal, and proposes links (depends_on edges or an outright "al
Multi-phase research workflow for building cited, audited knowledge bases.
21 skills · 10 agents
Domain-driven build workflow: vision grounding, EventStorming, bounded-context mapping, and a dependency-ordered task backlog driven to done by TDD workers.
12 skills · 9 agents
Design-system authoring workflow: reference-grounded Socratic foundation, themable token sets (light+dark), and pure HTML/CSS Tailwind component kitchen-sinks with accessibility and i18n baked in.
12 skills · 6 agents
Authoring workflow for Synaptic interactive-learning tracks: grounded ingestion, prerequisite-DAG structuring, node/question drafting, gap auditing, and CLI integrity gating.
9 skills · 6 agents
Cross-workflow utilities: /commit, /pr, /audit-context, /deckset.
8 skills · 5 agents
Bootstraps and maintains an AI-consumable map of an existing codebase with per-module CODEBASE.md and derived CLAUDE.md rules.
5 skills · 6 agents
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.