By krixon
Engineering workflow skills: planning, issue triage, audits, TDD, diagnosis, and skill authoring.
Land approved pull requests — merge each approved, bot-owned PR that is ready to merge, strip its issue's in-progress label, then tear down the local worktree and branch. Human-invoked only. Use when the maintainer has approved a PR and wants it merged and cleaned up, says "land it" / "land the approved PRs" / "merge and clean up", or just approved a PR review.
Sweep workflow state for staleness and clean it up, one human-confirmed action at a time — abandoned claims, quiet needs-info issues, orphaned worktrees and branches, and epics whose children have all closed. Human-invoked only. Use when the maintainer wants to tidy up stale workflow state, reap abandoned claims, clear orphaned worktrees, says "reap" / "clean up the workflow" / "sweep for staleness", or notices a crashed run left an issue stuck in-progress.
Audit a codebase for high-risk untested paths and surface them as findings. Static-first, no instrumented coverage run required. Use when the user wants to find test gaps, audit coverage, find untested critical paths, or asks "what's not tested?".
Sweep a codebase for dead code — unreachable branches, never-called functions, orphaned modules — and surface each as a finding. Static-first, no tool run required. Use when the user wants to find dead code, hunt unreachable or unused code, find orphaned modules, or asks "what's unused?" / "what can we delete?".
Harvest in-code debt markers (TODO/FIXME/HACK/XXX and known shortcuts) from a codebase and surface each as a finding, clustered by area. Static-first, no debt tool required. Use when the user wants to harvest TODOs, find tech debt, audit shortcuts, or asks "what's left to clean up?".
Sweep a codebase for unhealthy third-party dependencies — outdated majors on load-bearing deps, versions with known advisories, abandoned upstreams, and license drift — and surface each as a finding. Static-first, reads manifests and lockfiles, no installs or tool runs required. Use when the user wants to audit dependencies, find outdated or vulnerable deps, check for abandoned upstreams, or asks "what are we depending on that's risky?".
Audit a codebase for documentation that has drifted out of sync with the code — stale vocabulary, violated decisions, README/behavior claims the code no longer backs — and surface each as a finding. Static-first, no tool run required. Use when the user wants to check the docs are still accurate, find stale docs, or asks "is the documentation still accurate?" / "what drifted from the docs?".
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Runs pre-commands
Contains inline bash commands via ! syntax
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 library of Claude Code agent skills for engineering workflows — planning, issue triage, audits, TDD, diagnosis, and skill authoring. Packaged as a Claude Code plugin.
From the GitHub repo:
/plugin marketplace add krixon/skills
/plugin install skills@krixon
Or install from a local checkout:
git clone [email protected]:krixon/skills.git ~/dev/skills
Then in Claude Code:
/plugin marketplace add ~/dev/skills
/plugin install skills@krixon
Skills are namespaced once installed: /skills:diagnose, /skills:tdd, etc.
The workflow skills drive a small code adapter under bin/ (ADR 0008) that shells out to a few tools. On the machine that runs the skills you need:
gitgh (the GitHub CLI), authenticatedEach command requires only the tools it actually uses — the worktree group is pure git, so it doesn't need gh. The adapter binaries ship executable — a marketplace install is a git clone, which reproduces the committed file mode, so no chmod and no install step is needed. When a prerequisite is absent the adapter stops on its first call with a named halt on stderr — a JSON envelope whose reason is adapter substrate check failed and whose missing list names each absent prerequisite — rather than a cryptic exec error.
Targets macOS/Linux.
Start from the task, not the skill. Each entry is the head of a chain — run the command and it hands you to the next hop. Trace a chain by following each skill's handover default hop (skills/HANDOVER.md).
| I want to… | Run | …and the path from there |
|---|---|---|
| Validate a problem before designing a solution | /discover — grills the problem itself: who it serves, its value vs cost-of-inaction, its non-goals, how success is known | discover → design → slice → ready issues on the tracker |
| Stress-test a plan or design | /design — challenges it against the project's domain model, sharpens terminology, offers an ADR where one is warranted | design → slice → ready issues on the tracker |
| Turn a plan or this conversation into issues | /slice | one agent-brief issue, or N independently-grabbable issues under a lean epic, each marked ready-for-agent (AFK) or ready-for-human (HITL) |
| Find architecture / refactoring opportunities | /deepen | candidates + report surfaced → design to design the chosen one |
| Find what's not tested | /audit-coverage | findings → capture → needs-triage |
| Run a security audit | /audit-security | findings → capture → needs-triage |
| Check the docs haven't drifted | /audit-docs | findings → capture → needs-triage |
| Find swallowed errors / check error handling | /audit-error-handling | findings → capture → needs-triage |
| Find dead or unused code | /audit-dead-code | findings → capture → needs-triage |
| Find resource leaks (unclosed handles) | /audit-resource-leak | findings → capture → needs-triage |
| Harvest TODOs / find tech debt | /audit-debt | findings → capture → needs-triage |
| Find critical paths running blind | /audit-observability | findings → capture → needs-triage |
| Find performance hazards on hot paths | /audit-performance | findings → capture → needs-triage |
| Audit third-party dependency health | /audit-deps | findings → capture → needs-triage |
| File findings or observations as issues | /capture | deduped needs-triage issues for a human to triage |
| Triage incoming issues | /triage | each routed to ready-for-agent, ready-for-human, needs-info, or wontfix |
| Start implementing a ready issue | /pickup | claims it → routes by kind to tdd / diagnose / write-skill / docs / config → review gate → opens a PR |
| Ship a small fix with no tracked issue | /patch | branch in a worktree → review gate → a no-issue PR (body led by No-issue:) a human lands; human-invoked, for fixes too small to file |
| Build a feature test-first | /tdd | red-green-refactor loop (usually reached via pickup) |
| Debug a hard bug or perf regression | /diagnose | reproduce → minimise → fix → regression test (usually reached via pickup) |
| Ask questions on a PR review (not just request changes) | Comment them on the PR, then /pickup | pickup's rework query catches the unresolved thread → routes the questions to field, where the agent works each to a converged answer with you → the answers post back to the thread |
| Merge an approved PR | /land | merges, strips in-progress, deletes the branch/worktree (human-invoked only) |
| Clean up stale workflow state | /reap | sweeps abandoned claims, quiet needs-info, orphaned worktrees/branches, and emptied epics → proposes each fix, confirms per item (human-invoked only) |
npx claudepluginhub krixon/skills --plugin skillsHarness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).