Claude Skills by alexmond

Given targeted work — a problem, a deliverable to improve, an idea to develop — assemble the right panel of role-specialized agents, pick a coordination style suited to the task, and run a generate-critique-refine loop until it converges. The skill picks the team and process from the work itself. Use when the user hands over work wanting multiple expert perspectives — when they say "brainstorm", "get a team on this", make it "better / more appealing / more convincing", explore options, or pre...

Use this to take a delivery task through a task-fit roster of role subagents — composed for the work and run as a gated relay (architect -> dev -> qa -> deployer is one example lineup; the roster is dynamic), where each role runs as a subagent on its own model tier. Trigger on "run the crew on", "ship this feature", "build and test X", "take this through the cycle", or any multi-step implementation/refactor/release task that benefits from separated design, build, verify, and deploy phases. Th...

Set up CLAUDE.md to maintain a living Decisions & Learnings log that stays bounded as the project evolves — a format contract enforced at write time, four pruning pressures, and a coverage check for the essentials. Use when the user says "make CLAUDE.md evolve", "self-updating CLAUDE.md", "add a learning mechanism to CLAUDE.md", "decisions log", "ADR-style notes", "how do we keep CLAUDE.md current", "compact CLAUDE.md", "CLAUDE.md is getting too big" — or when a CLAUDE.md exists with no mecha...

Drive a GitHub issue from branch to pull request — read the issue, create a feature branch, implement the change, run the project's build and tests, then commit, push, and open a PR. Language and build-tool agnostic. Use when asked to implement, fix, or resolve a GitHub issue by number.

Save a learning to project memory. Invoke automatically (without the user asking) whenever a task required more than one fix cycle to resolve — e.g. a test failed and needed a second attempt, a compile error required a correction, an API behaved unexpectedly, or an assumption proved wrong mid-task. Also invoke when the user explicitly asks to remember something — "remember this", "save this learning", "don't make that mistake again". Do NOT invoke for routine single-pass work.

Coding standards and formatting for Maven/Java projects using spring-javaformat and Checkstyle. Use when the user says "fix the code-style violations", "reformat this module", "the build is failing on Checkstyle", or asks what the project's formatting and naming rules are. For PMD static analysis specifically, see the `pmd` skill.

Check JaCoCo code coverage for a Maven/Java project against a gate, then drill into per-module, per-class, and per-line gaps to target tests. Use when the user says "what's our test coverage", "check coverage", "which classes need tests", "are we above the coverage gate", or is planning tests to reach a threshold.

Set up, run, and triage PMD static analysis on a Maven/Java project — the ruleset, the plugin wiring, parsing the report by rule and priority, copy-paste detection, and how to suppress a finding at the right level. Use when the user says "run PMD", "PMD is failing", "fix the PMD violations", "add static analysis", "what's PMD complaining about", "suppress this rule", or when a Maven build fails at the validate phase with rule violations.

Run pre-commit quality checks for a Maven/Java project — auto-format, validate Checkstyle and PMD, and run the tests. Use before creating a commit, and when the user says "run the pre-commit checks", "am I ready to commit", "check this before I commit", or "is the working tree clean".

Keep agent-written memory files from rotting. Audits ~/.claude/projects/<slug>/memory/ for facts the tree now contradicts — stale version pins, passed sequence facts, vanished paths — plus MEMORY.md index drift, and enforces the memory format contract at write time. Use when the user says "audit my memory", "memory is stale", "clean up agent memory", when a session recalls something that turns out false, or proactively when the SessionStart banner reports rot candidates.

Think on a canvas, then compile the map into an organized prompt. Idea nodes connect freely and a deterministic compiler turns the whole map into structured Markdown; ✦ expands any node with `claude -p` running in the repo, so ideas stay grounded in the project. Maps save as JSON Canvas, so they are git-diffable and reworkable later. Use when the user says "mind map", "map this out", "brain dump", "let me sketch this first", "turn my notes into a prompt", "expand this idea", "flesh this out",...

One visible channel for every long-running process — local sweeps, backgrounded commands, and work handed to external systems (a CI run, a media-server queue, a long download). The server that serves the live progress page IS the tracker — a stdlib Python daemon holding live jobs in memory, auto-spawned by the first producer — with learned ETAs, pre-start duration forecasts, and stall/orphan detection from each job's own history. Use when starting any operation expected to exceed ~10 seconds,...

A hook-driven coach that reads every prompt sent to Claude Code and rewrites it toward proven prompting habits — definition-of-done, scoped references, guardrails, verification. Rules graduate as they are demonstrated, so the coaching fades as the user improves. The hook runs on its own, but load this skill when the user says "coach pause 10", "coach off <rule-id>", "coach on <rule-id>", "coach reactivate no-definition-of-done here", "disable praise", "coach ack every 10", "coach that was wro...

The "discover" orchestrator on the shared role substrate — counterpart to dev-crew (deliver) and brainstorm-panel (decide). Composes a task-fit team of coverage roles, partitions the space into non-overlapping angles, fans them out as parallel agents under one output contract, then dedups and adversarially verifies the merged, cited findings. Use when the user asks for something "comprehensive", "exhaustive", "as many as possible", "every X", or a "catalog/survey/landscape of Y" — any gatheri...

Design a scalable system — architecture, data flow, APIs, schema, caching — and implement its minimal production version, applying the repo's evolving `architect` role (.claude/roles/architect.md). Use when the user says "/roles:as architect", "design the architecture for X", "how should this system be structured", or "act as the architect".

Run any registered role from the repo's evolving role repository (.claude/roles/) against a target — the generic solo dispatcher for roles without a wrapper skill. Use when the user says "/roles:as <role>", "run the <role> role on X", "act as the <role>", or names a registry role that has no skill of its own.

Scaffold a complete, production-ready application — sound architecture and a minimal but scalable MVP — applying the repo's evolving `builder` role (.claude/roles/builder.md). Use when the user says "/roles:as builder", "build X from scratch", "scaffold a new service", or "stand up a new app".

Investigate a bug to its true root cause and ship a robust production-ready fix, applying the repo's evolving `debugger` role (.claude/roles/debugger.md). Use when the user says "/roles:as debugger", "act as the debugger", "find the root cause of X", or hands over a stubborn failure that has resisted a quick fix.

Audit + refine the role files in `.claude/roles/` so personas keep improving after first graduation: consolidation candidates (3+ Learnings on one topic), staleness (Learnings citing vanished files/symbols), Body drift, and solo-to-core graduation — all proposals user-gated, never silent rewrites. Use when the user says "evolve roles", "audit roles", "consolidate role learnings", "refresh roles", "what should we graduate", or "roles are getting stale"; also periodically when many sessions hav...

Find bottlenecks and unnecessary work, and improve speed, memory, and scalability with measured, behavior-preserving changes, applying the repo's evolving `optimizer` role (.claude/roles/optimizer.md). Use when the user says "/roles:as optimizer", "optimize X", "why is this slow", or "reduce the memory/cost of Y".

Restructure code into clean architecture — separate concerns, raise modularity, reduce coupling, behavior unchanged — applying the repo's evolving `refactorer` role (.claude/roles/refactorer.md). Use when the user says "/roles:as refactorer", "refactor X", "untangle this module", or "clean this up without changing behavior".

Map an unfamiliar or large codebase and improve it safely — behavior unchanged, quality raised — applying the repo's evolving `reviewer` role (.claude/roles/reviewer.md). Use when the user says "/roles:as reviewer", "review this codebase", "walk me through this repo", or "assess the quality of X".

Pressure-test a plan, finding, or design — steelman it, attack correctness/necessity/cost, and deliver a REFUTED / HOLDS / HOLDS-WITH-CUTS verdict — applying the repo's evolving `skeptic` role (.claude/roles/skeptic.md). Use when the user says "/roles:as skeptic", "pressure-test this", "poke holes in this plan", or "is this actually sound".

Use whenever a screenshot is captured or supplied — a bug report, a Playwright capture, a design review, a "does this look right". Read the WHOLE image, not just the thing it was sent for: sweep it against the defect checklist below, file the findings, and add them to the todo list. When a defect is later reported that this sweep should have caught, append the miss to the Misses log (in the consuming repo's .claude/screenshot-sweep/ when installed as a plugin) and add the check that would hav...

Generate a presentation-worthy screenshot deck of the current product — discovers the aspects worth showing (the CLI's `--help`, the web app's routes, the library's quickstart), captures each with the *right tool per surface* (project's existing browser driver for web — Selenium-Java if none / Playwright for Node-Python / VHS for CLI / Freeze for code snippets), and assembles a captioned, narrative-ordered deck under `presentation/` ready to walk through in a meeting, paste into slides, or se...

Defensively scan a codebase (or a specific path) for OWASP-style vulnerabilities — injection, path traversal, unsafe reflection/deserialization, SSRF, hardcoded secrets, and XXE — and report each confirmed finding with its file:line location, severity, and concrete remediation. Use when asked to do a security audit, review code for vulnerabilities, find injection/SSRF/secret-exposure risks, or harden a project before release.

Audit SKILL.md files against published skill-authoring guidance — frontmatter validity, whether the description actually says when to trigger, body size, progressive disclosure, and broken references. Use when the user says "lint my skills", "check my SKILL.md", "review this skill", "do my skills conform", "audit the marketplace", or is about to publish, rename, or graduate a skill. Use proactively right after writing or editing any SKILL.md, and whenever a skill turns out not to trigger when...

Spring Batch 5/6 development — job repository wiring, restart versus a new JobInstance, flow and deciders, tasklet transaction boundaries, partitioning strategy, and testing. Use when writing or debugging a Spring Batch job, when a job "completes" having done nothing, when asking whether a job resumes after a restart, when choosing between a tasklet and a chunk step, when partitioning work across workers, or when a batch test passes and pins nothing. Grounded in runnable samples in spring-boo...

Discipline for bug fixes: treat the local fault as one instance of a class — check scope (grep the pattern elsewhere in repo-calibrated order), name the class, propose prevention from the repo's actual lint/pre-commit/test tooling, and report findings before any patch (default: local fix only). Use whenever a bug is found during ordinary work, or on "is this systemic?", "look wider", "find related bugs", "fix the class, not the instance", or `/systemic-fix` (`recalibrate` redoes the ~30s firs...

Review the open tickets, rank them, work out what is genuinely startable versus blocked, and run the startable ones in parallel through to merged — then KEEP GOING: re-triage every time an agent finishes, so the backlog drains without being re-asked. Use whenever the ask is "what's next", "review tickets", "triage the backlog", "prioritise", "continue with priorities", "start in parallel", "drain the backlog", or a bare "go" against the repo's backlog, and stay in the loop until the queue is ...

Audit and tighten the current repository so Claude Code works faster and more accurately in it — verify the existing CLAUDE.md still matches reality, tighten the build/test/lint verification loop, add only the missing static guardrails, and reduce permission friction. Use when the user wants to "tune", "audit", "harden", or "onboard" a repo for Claude/AI agents, fix a stale or thin CLAUDE.md, or make agent runs more reliable. Works in any repo; pass "audit" to report without changing anything.