Engineering workflow skills, review agents, and a /feature pipeline: orientation, planning, test-driven development, systematic debugging, refactoring, error handling, observability, containers, CI pipelines, dependency upgrades, migration planning, incident postmortems, code review, completion verification, and performance work.
22 skills · 2 commands · 7 agents
The research lifecycle for a master's thesis or paper: research question and proposal, multi-source literature search, structured paper notes, a verified BibTeX bibliography, citation verification with a hook that catches invented keys as they are written, argument-first literature synthesis, research design and design science, statistics planning and reporting, PRISMA systematic reviews, citation-graph expansion, reproducible experiments and benchmark reporting, section-by-section academic writing, LaTeX and Word toolchains, publication figures, an examiner-style reviewer and a paper critic, proofreading, supervisor updates, a research log, and progress tracking.
21 skills · 3 commands · 2 agents · hooks
Ship a Next.js app or an Express and Prisma API to a VPS under Coolify, end to end: runtime-only images with health probes and fast shutdown, one GitHub Actions pipeline from pull request to production through GHCR, the Coolify application and API token, a host ops toolkit with alerting health checks and verified backups, server-side deploy verification, a zero-downtime domain cutover, and a read-only deployment reviewer.
9 skills · 3 commands · 1 agent
Version-control skills for the whole branch lifecycle: committing, worktrees, merge conflicts, undo/recovery, PR descriptions, changelogs, releases, and branch cleanup.
9 skills · 3 commands
Session and personal productivity: collaborative brainstorming, plan stress-testing, session handoffs, meeting notes, a decision log, learning plans, a /prime command that loads project context, and a /weekly-review that assembles the week from evidence.
8 skills · 3 commands
Pull-request review as a set of specialists: a read-only code explorer, a behaviour-preserving simplifier, hunters for silent failures and test gaps, a type-design reviewer, and a /review-pr command that runs them in parallel and merges one ranked review.
1 command · 5 agents
Documentation that stays true to the code: architecture decision records with an /adr command, READMEs verified against the repository, developer guides in the four documentation types, and OpenAPI specifications kept in sync in CI.
4 skills · 1 command
Security review toolkit: OWASP-aligned code review, dependency and secret auditing skills, a security-auditor agent, and a full-codebase /security-audit command.
3 skills · 1 command · 1 agent
Investigation toolkit: a deep-research subagent for multi-source work with citations, plus name-clearing skills for software projects (registries, app stores) and for businesses (social handles, storefronts, company registers).
2 skills · 1 agent
Testing toolkit: browser-based end-to-end verification with Playwright and a /write-tests command that generates a suite matching project conventions.
2 skills · 1 command
Prose for a human reader: strip machine-writing tells from any text, draft emails that get answered, and outline talks and defences slide by slide.
3 skills
Automation hook that runs the project's own formatter (Biome, Prettier, gofmt, rustfmt, or ruff) on every file Claude edits, so changes land already formatted.
hooks
Desktop notifications for Claude Code: surfaces permission requests and attention prompts as native macOS/Linux notifications so long sessions can run in the background.
hooks
Context-preservation hook that snapshots the full session transcript before every compaction, keeping the newest ten snapshots per project.
hooks
Deterministic guardrails, active immediately after install: a compound-command deny list, sensitive-file protection, and prompt-injection screening.
hooks
Observability hook that appends one sanitized JSON line per tool call to a local log, with secret redaction and payload truncation.
hooks
Session workflow guardrails, active on install: a session-start situation report, a stop gate that runs the tests before Claude finishes, type errors fed back after each edit, git footgun protection, a branch-first nudge, and a subagent audit log.
hooks