Coding & Development

Provides access guide for Huawei Cloud Graph Database GES service. Covers Cypher queries, GQL queries, schema/label management, summary info queries, graph data editing and more. Use this skill when users want to operate Huawei Cloud graph database GES service via terminal.

Validate that a Figma icon component is organized correctly before exporting it from the Koobiq icons file (figma:sync / figma-export). Checks node type, name format, size, layout, the root shape/shape-2 structure, scale constraints, fill color contract, and the description metadata — and verifies size variants of the same icon agree with each other. Use this whenever the user wants to check, review, validate, or "проверить" a newly drawn or unpublished icon component against the library conventions, or asks whether an icon is ready to export — even if they only paste a Figma link or an icon name. Run it before figma:sync when adding or editing icons.

Configure Huawei Cloud OBS static website hosting with Python SDK and a custom domain. Use when the user needs to enable or repair OBS website hosting, set index or error pages, expose an existing bucket for public website access through a custom domain, or connect that domain through Huawei Cloud DNS when Huawei manages the zone. Triggers include OBS static website hosting, website endpoint, index page, error page, public-read bucket website access, custom domain CNAME, Huawei Cloud DNS, setBucketWebsite. 中文触发词包括：OBS 静态网站托管、网站托管、自定义域名解析。

Round-based interview that settles every open decision with a human before implementation starts and ends in a decision record; triggers on "grill me" and "frag mich aus".

Upload a local file or directory to a specified Huawei Cloud OBS (Object Storage Service) bucket with obsutil, and set up a scheduled/periodic upload task (crontab) that syncs a local directory to an OBS bucket. Also lists OBS buckets with storage capacity and object count. Use this skill when the user wants to: (1) upload a specific local file or directory to a named OBS bucket, (2) set up a scheduled/periodic upload task that automatically syncs a local directory to a named OBS bucket, (3) list their OBS buckets and check capacity and object count. Triggers include: user mentions "upload to OBS bucket", "上传文件到OBS", "上传目录到OBS", "定时上传到OBS", "目录同步到OBS桶", "OBS scheduled upload", "OBS周期上传", "OBS bucket capacity check", "OBS桶容量查询". Generic phrases such as "object storage", "upload file" or "upload directory" alone are NOT sufficient to trigger this skill.

Add semantics (what the data means) to an existing Croissant metadata file for a biomedical knowledge base — group columns into typed entities (genes, diseases, ...), type rows as Biolink associations, mark identifier prefixes, units and code values — using only Croissant 1.1 constructs, and check the result against the real data. Use when asked to add semantics, Biolink types, identifier prefixes or entity/association structure to a croissant.json.

Bring the branch up to date with its base — rebase when nothing has been published, merge when it has, resolve the mechanical conflicts, and stop at G5 with both sides shown for a semantic one. Use when the maintainer invokes $ao [from <base>], asks to pull main in, catch the branch up, or rebase onto main, or whenever $aka, $murasaki or $en needs the branch current. Kurapika classifies every conflicted path before touching one and never picks a side on a conflict of meaning. Never pushes, never force-pushes, never rewrites a commit that is already on the remote.

Run each request in an independent effort worktree by default, rechecking isolation on continuations. Warm up, author, verify and commit locally, handle launch, report and ring the bell, then wait. Use on $ren <request> or ordinary work in a repository using the turn loop. Ren composes six skills; it never pushes or opens a PR and ends only on the maintainer's $aka or $tensho call.

Run an assumptions-first architecture design and ship it as an interactive design doc. Ground truths land in a register with stable IDs, every design fork goes through a question round with a recorded escape hatch, an adversarial review attacks the middle draft, and every estimate is gated on a named spike. The deliverable is a registers.json-driven single-file HTML doc that opens on a hand-written executive summary, cites every entry by a human-readable handle, carries declared interactive components, and answers questions in the page when an AI config is deployed beside it, plus a generated PDF, written as a humble proposal that explains and asks for feedback. Use when asked to "write a design doc", "architecture proposal", "help me design <system>", "redesign <system>", "assumptions-first design", or to turn a systems discussion into a reviewable design document.

Use Cruma to put local things on a public HTTPS URL and manage them - a dev server port, a folder of files, several named sites at once, password, sign-in or IP protection, agent-hosted dev servers. Also for inspecting or changing the user's own running Cruma desktop or headless agent (sites, routes, processes, logs), and for installing Cruma. Anything beyond a plain share is looked up in Cruma's own docs first. Asks before exposing anything.

Implement a Redocly CLI lint rule in a user's redocly.yaml or custom plugin for an OpenAPI, AsyncAPI, or Arazzo description. Use when the user wants a convention enforced, asks for a configurable rule or a plugin rule, debugs a rule that never reports or gives false positives, or migrates rules from another linter such as Spectral.

Review an existing test case document (any sheet/CSV/markdown) against the acceptance criteria (AC), edge cases (EC) and business rules on its governing ticket. Build a traceability matrix, judge coverage and row quality, and report Good vs Need Improve, each anchored to a quoted source. Always notify the finished review to the ticket project's own TC-review channel, asking only which thread. Use when the user asks to review test cases or a TC sheet, check TC quality or coverage before execution, audit someone else's test case document, compare TCs against AC/EC, or wants feedback on a TC document without executing it (e.g. "รีวิวเทสเคส"). Also /tc-review. Also when a TC sheet and ticket are pasted with "is this OK/complete?". Do NOT use for drafting new test cases (tc-fe-prep-workflow / tc-api-prep-workflow), executing a ticket (testing-ticket-workflow), reviewing recorded evidence clips (review-result-workflow), or retesting a fix (retest-bug-workflow).

Every Zammad ticket, article, and Knowledge Base operation as one CLI and MCP server - plus a team-management layer (agent load, customer health, aging backlog, escalation triage, churn risk, feedback mining) the Zammad API can't answer in a single call. Trigger phrases: `who is overloaded in support`, `which customers are at risk`, `what tickets are open too long`, `who should we escalate`, `what feedback are we getting`, `search zammad tickets`, `use zammad`, `run zammad`.

Use when managing Huawei Cloud ECS (Elastic Cloud Server / 弹性云服务器) instances — lifecycle management (create / start / stop / restart / delete), day-to-day change operations (resize / reinstall OS / reset password / replace keypair / bind & unbind EIP / change security group / add & remove NIC / attach & detach volume), instance querying, quota/flavor/image inspection, mandatory async-job polling after every write, and multi-step root-cause diagnosis of ECS instance creation failures (quota → flavor → image → network → keypair → disk). Provides 23 huawei_* actions: huawei_list_ecs_instances, huawei_get_ecs_instance, huawei_list_ecs_flavors, huawei_list_ecs_images, huawei_list_ecs_quotas, huawei_diagnose_ecs_create_failure, huawei_analyze_ecs_health, huawei_create_ecs_instance, huawei_start_ecs_instance, huawei_stop_ecs_instance, huawei_restart_ecs_instance, huawei_delete_ecs_instance, huawei_resize_ecs_instance, huawei_reinstall_ecs_os, huawei_reset_ecs_password, huawei_replace_ecs_keypair, huawei_bind_ecs_eip

Complete guide to AAD, CAD for agents. Covers everything AAD can do - every MCP tool and argument, model.py, gears, selectors, spec.yaml, process checks, renders, drawings, the viewer, import and export, the library, the CLI, remote mode - and what does not work yet. Use whenever a user wants a mechanical part, an assembly, a drawing or a CAD file, or asks what AAD can do.

Work the To Do list on one of the user's self-hosted Slipdock boards unattended — take the top card, do it, descend into its subcards, move it through in progress to done, and leave the record on the card. Use when asked to work through the To Do list, the board or a backlog without being driven card by card ("pick up whatever is next and do it", "keep working the board", "clear the To Do list"), and whenever this is being run repeatedly from /loop, a schedule or a cron — including passes told to take their work from a runner pool's job queue.

Every nemia verb you run, by who answers for it — a read, yours unasked, one that prompts the person, one that answers what waits on them — with its usage and help.

Decompose a spec's PRD/TechSpec into a dependency-ordered task graph — vertical-slice task files plus a machine-parseable _tasks.md DAG manifest under docs/specs/<slug>/, including the authored QA gate decision, decided autonomously from the Spec and escalated only when authority, architecture, or blast radius is at stake.

Coordinate explicitly requested Wollipog child-session issue campaigns through merge, cleanup, recursive follow-ups, and archival. Use only when the user invokes this skill or explicitly asks to orchestrate or delegate issue implementation across sessions. A request to claim, implement, or fix multiple issues alone stays in the current session and does not trigger this skill. Do not trigger merely because the issues concern Orchestrator features.

DWS cluster I/O overload root cause diagnosis skill, based on KooCLI v3.2.0+ and DWS Autopilot MCP Server. Automatically collects I/O metrics, analyzes root causes (customer-side / system-side) via three-stage decision tree, and outputs a standardized diagnosis report. Applicable to DWS cluster I/O usage too high, I/O alarms, disk I/O load anomaly scenarios. Trigger words: "I/O高", "I/O告警", "I/O诊断", "I/O过载", "IO高", "IO告警", "IO诊断", "IO过载", "磁盘IO负载异常", "util打满", "await高", "high I/O", "I/O alarm", "I/O diagnosis"

Runs the package.json check script (monorepo-aware) and auto-fixes until green, escalating audit findings, whenever work must prove runnability; triggers on "pnpm run check".

Every PRTG API v2 endpoint, plus a fleet view across every client's PRTG core: alarms, status, handoffs, hygiene and safe bulk actions in one command. Trigger phrases: `what's down across all our PRTG cores`, `acknowledge this alarm on every client`, `pause monitoring for patch night`, `which PRTG cores are behind on version`, `use prtg`, `run prtg`.

Send the branch out — gyo (lint) before rewriting history, squash only unpushed commits, bring the base underneath them, gyo again if catch-up changed the tree, and push. The whole skill runs only on the maintainer's publish call. It never runs project-wide tests or coverage; those belong to mukai through kotoamatsukami and byakugan.

Drive one open pull request to verified readiness under a mandatory acting-cycle cap. Use when $mukai hands over its PR, or when the maintainer invokes $en on <CODE>#<N> for an open PR. Pending CI or review keeps the run active; quiet polls spend no cycle. At Ready, merge only a recorded Futon-owned stacked delivery through Nen's bounded guard; hand every other PR to the maintainer. Never cast a review vote. After this run completes, the caller offers the wrap-up — $cleanup, required, with $third-hand optional beside it — En does not own it.