engineering

Plugin engineering from marketplace kzarzycki-agent-skills in kzarzycki/agent-skills. Commands read from the repo's manifest — today.

Skills (34)

• ask-matt

  Router over this package's skills. Use when the user asks which skill or flow fits their situation, or what to do at a phase boundary.

• claude-handoff

  Hand the current conversation off to a fresh Claude Code background agent that picks up the work immediately.

• code-review

  Review the changes since a fixed point (commit, branch, tag, or merge-base) on two separate axes, the repo's coding standards and the originating spec, reported side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to \"review since X\".

• codebase-design

  Shared vocabulary and principles for designing deep modules. Use when designing or improving a module's interface, finding deepening opportunities, deciding where a seam goes, making code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

• diagnosing-bugs

  Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose" or "debug this", or reports something broken, throwing, failing or slow.

• domain-modeling

  Build and sharpen a project's domain model, its GLOSSARY.md glossary and its ADRs. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.

• grill-with-docs

  Grill a plan or design while recording its glossary (GLOSSARY.md) and ADRs as decisions land.

• grilling

  Interview the user in rounds of questions until every decision in a plan, design, or idea is settled. Use when the user wants to stress-test their thinking or uses a "grill" phrase such as "grill me".

• handoff

  Write the current conversation into a handoff document for a fresh agent to pick up, in another session, harness or directory.

• implement-spec

  Implement a spec and its tickets on one integration branch, working the ticket graph's frontier concurrently.

• implement

  Implement a piece of work from a spec or tickets, test-first, then review and commit it.

• improve-codebase-architecture

  Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

• loop-me

  Grill the user into implementable specs for the recurring workflows they want to build or delegate, kept in this workspace.

• pr

  Write a PR body as a Summary visual of the change, before-and-after Evidence, and its Merge Danger. Use when writing a PR body.

• prototype

  Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

• research

  Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.

• retro

  Run a retrospective on a coding session and propose changes to the agent's environment (checks, coding standards, steering files, tooling, information access) that would improve future runs.

• tdd

  Test-driven development in red-green vertical slices at agreed seams. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

• teach

  Teach the user a topic over several sessions, using the current directory as a stateful teaching workspace.

• to-spec

  Synthesize the current conversation into a spec and publish it to the project issue tracker, without interviewing the user.

• to-tickets

  Break a plan, spec, or the current conversation into tracer-bullet tickets that declare their blocking edges, and publish them to the configured tracker.

• triage

  Move issues, and external PRs where the tracker allows, through triage roles - categorise, verify, grill if needed, and write agent-ready briefs.

• wait-what

  Re-pitch a message that did not land: add the missing context, in Simplified Technical English and the repo's GLOSSARY.md vocabulary.

• wayfinder

  Chart an effort too big and too foggy for one agent session as a shared map of decision tickets on the issue tracker, then resolve them one per session until the way to the destination is clear. Use to start or continue a wayfinder map.

• wizard

  Generate an interactive bash wizard that walks a human through steps only they can perform, such as provisioning infrastructure, setting up credentials or CI secrets, clicking through an unfamiliar third-party dashboard, or running a one-off migration or cutover. Not for steps the agent can perform

• agent-hygiene

  Keep a dev account where coding agents work clear of what rebuilds itself (merged worktrees, tool caches, idle environments, dead Omnigent terminals) with the agent-hygiene script, settle the routine cases it keeps, and pass on only what needs the owner. Use for a scheduled hygiene round, or when an

• audit-third-party-software

  Audit third-party code (repos, npm packages, binaries, plugins) for telemetry, exfiltration, supply-chain risk, and hostile defaults before install; verdict SAFE/CAUTION/UNSAFE with file:line evidence.

• chief-of-staff

  Act as a project's chief of staff, the one agent the owner talks to while leads do the work. It sorts each owner message, starts one lead per ask as a top-level session, relays follow-ups with the owner's words quoted, and watches the leads to surface their results, questions and stalls. Use only wh

• context-extractor

  Draft or extend a project's CLAUDE.md or AGENTS.md with the commands, conventions and gotchas an agent cannot infer, for any project type. Use on "extract conventions", "analyze project patterns", "generate CLAUDE.md", "what are the conventions", or /conventions.

• demo

  Record a short video or take a screenshot of a change running and post it to its PR, so a reviewer judges a visible change without checking out the branch. Use when a PR changes what a user sees (a command and its output, a page, a screen, a speed-up), on "record a demo", or when the project's instr

• engineering-loop

  The engineering loop, run for every change that will land as a PR, however small - intent, a spec the owner reads, build, gates, proof, an independent verifier, triage, the pass cap and autonomous landing. Use from the owner's request onward, when working or verifying part of one, or when deciding w

• omnigent-telemetry

  Report Omnigent spend, tokens by model, session trees, scheduled-task health and Codex rate limits over the last N days, with one-line anomaly flags and an optional self-contained HTML page. Use for a retro across Omnigent sessions, or when asked where spend or a quota went.

• operating-omnigent

  Operate a local Omnigent install — crashes, blank UI, restarts, session attach/resume/dispose, process topology — and its raw sys_* MCP tools. Starting and collecting agents through a backend is orchestrating-agents'.

• orchestrating-agents

  Start, message, wait on, watch, list and close agent sessions through one backend (native subagents, Omnigent or herdr-link), with reports as files and questions as files. Use when starting a child agent such as a worker, verifier or lead, waiting on or collecting the agents you started, or choosing

Similar plugins