jcordon5

Expert guidance for Apple's `container` CLI. the open-source tool for running, building, and managing Linux containers and persistent Linux "container machines" as lightweight VMs on Apple silicon Macs. Use when the user wants to run containers or Linux dev environments on a Mac with Apple `container`, replace Docker Desktop, Colima, Lima, or OrbStack, use commands like `container run`, `container build`, `container machine create`, or `container system start`, or configure ports, volumes, networks, registries, DNS, or container networking on macOS. Also use for diagnosing Apple `container` runtime, VM, port, DNS, or networking issues. Do not use for Docker Desktop, Podman, Kubernetes-specific workflows, or generic Dockerfile authoring unrelated to Apple `container` on Mac.

Persistent knowledge wiki and memory system that compounds across conversations using an Obsidian-compatible vault. Consult at the START of any non-trivial task to recall project context, preferences, and past discoveries. Ingest sources and build wiki pages when accumulating knowledge on a topic. Save at the END when new knowledge was gained. Use when: starting work on a known project, dealing with auth/credentials/config, making stylistic choices, user corrects output, user says "as always"/"remember this"/"save this"/"don't make me repeat". Also trigger for: "ingest this", "add to my wiki", "what do I know about X", "research notes", "lint my wiki", deep research sessions, reading notes, competitive analysis, or any time the user wants knowledge to accumulate and compound across multiple sessions. Do NOT use for trivial one-line questions.

Safe, conversational disk cleanup assistant for macOS and Linux. Use this skill whenever the user wants to free up disk space, recover storage, or figure out what is filling their drive — phrases like "my disk is full", "clean up my mac", "I'm out of space", "what's taking up all my storage", "free up some gigs", "optimize disk usage", or "why is my startup disk almost full". Use it even when the user doesn't say the word "clean": if they mention low disk space, a full drive, large files/caches they want gone, or ask where their storage went, reach for this skill. It scans adaptively, classifies everything by safety, shows a local visual dashboard, asks before deleting anything, and reports how much was recovered. Do NOT use it for cloud storage quotas, deleting a single known file the user already pointed at, or database/disk-partition administration.

Lets Claude reach the user on Telegram during long or unattended runs — to send progress updates and, crucially, to ask BLOCKING questions and WAIT for the user's reply before continuing. Use this skill whenever you're working autonomously toward a goal for a while and the user may be away from the computer, OR whenever you hit a decision/confirmation/missing-context that blocks progress and you'd otherwise stall or guess. Triggers on: "ping me on Telegram", "message me when…", "ask me if you get stuck while I'm out", "let me know the progress", "work on this for a few hours and tell me if you need anything", or any setup where you'll run unattended and need a human in the loop. Also use it proactively (even if not asked) when a long task you're already running needs a human decision and nobody is watching the terminal. If it isn't configured yet, walk the user through the one-time setup here.

Automatically patches skill files (SKILL.md and associated scripts) when Claude detects and resolves an error during skill execution. Use this skill whenever Claude has just worked around a problem caused by a skill's instructions — such as a wrong path, an unavailable command, an outdated flag, a broken script, or a missing dependency — and wants to record the fix so the same error never happens again. Triggers automatically after error recovery during any skill usage. Also triggers when the user says things like "fix this skill", "update the skill", "save this fix", "patch the skill", or "don't make me deal with this error again".