Afegits recentment
Els últims skills afegits al catàleg. Descobreix les novetats!
conductor
medTech-group
Run an unattended build run - ask the owner the run charter's questions once, then hand each ready ticket to its own cloud session, check in on a schedule, keep PRs moving to merge, release as the charter allows, and write the final report. Also resumes or stops a run. Use for "run the backlog unattended", "work over the weekend", "conductor resume", "stop the run".
build-ticket
medTech-group
Build one ready ticket from claim to merge - tdd at the spec's seams, the review panel in rounds until clean, a pull request with its findings table, CI green, the local suite's evidence, and the merge under the merge rule. Use for "implement #N", "build #N", "build #N with a light review" ("quick build #N"), or a ticket session's briefing.
start
medTech-group
The guided start - ask what we're working on with clickable choices, split a request into items, propose a path for each (a quick ticket, finding a bug's cause, a spec, a map), record the plan as issues, and offer every next step as a choice, so the person never has to remember a command. Use at the first message of an attended session, on /start, or when the person isn't sure what to do next, and when they say "explain less" or "explain more".
setup-workflow
medTech-group
Set a project up for the pairnexus workflow in solo mode - explore it, ask only what attended work needs (experience, sensitive data, the explorer, the mode, the owner), read the rulesets back, write the profile, vendor the kit for cloud sessions, write the Matt Pocock skills' per-repo docs, labels, issue templates, the pull request template and CODEOWNERS, and open one pull request to merge and start working. Run once per project.
run-report
medTech-group
Write the final report of an unattended run (or a progress report mid-run) - what merged, what was set aside and why, what's live, the outward changes to confirm or undo, the decisions for the owner, the by-hand checks and the follow-ups. Used by the conductor at the end of a run, or alone.
next
medTech-group
Find the next ready piece of work (a map decision, a spec to write, or a build ticket) and do it under the workflow's flow. Optional argument - "map", "spec", "build", or an issue number.
merge-check
medTech-group
Answer "is #12 ready to merge?", "can I merge this?" or "is my change ready?" in the chat, from workflow merge-check --json. "Not yet" gives each reason in plain words and offers the fixes as choices ("Review it now", "Bring it up to date", "Resolve the conflict", "Ask Claude to fix it"); "Ready" gives the try-it link first, then "Merge it now", which checks again on the same head commit and merges with a merge commit pinned to it. Posts nothing on GitHub. Use when someone asks whether a pull request is ready or can be merged, or picks "Check whether #N is ready to merge".
conductor
medTech-group
Run an unattended build run - ask the owner the run charter's questions once, then hand each ready ticket to its own cloud session, check in on a schedule, keep PRs moving to merge, release as the charter allows, and write the final report. Also resumes or stops a run. Use for "run the backlog unattended", "work over the weekend", "conductor resume", "stop the run".
check-setup
medTech-group
Check the set-up - gather what only a Claude session can see (the rulesets and repository settings through the GitHub tools, the base branch's latest check runs and statuses, the open pull requests, the cloud environments, the workflow:run issues, Claude Code's version), hand it to workflow doctor --json as files, show the report worded by the person's experience with the most serious first, and offer each upgrade it names as a choice with "Not now". Changes nothing itself. Use when the person asks to "check the set-up" (or "check the workflow set-up", "is my project set up properly?"), or picks it where the guided start offers it.
ai-worker-ui
hoang-phan
Applies ai-worker's CRUD-screen discipline. Use for any task touching app/controllers/, app/views/, or config/routes.rb — the Projects, Workflows (including runbooks and their Stages), PromptTemplates, Jira/Bugsnag syncs and Settings screens.
ai-worker-stage-types
hoang-phan
Checklist for adding, renaming or changing an ai-worker stage type, a Workflow status, or an agent result marker line (e.g. LOCAL_TEST_RESULT). Use when a change spans Stage, StageRun::ACTIONS, Workflow, StageExecutor, prompts, AiCli::Output and the scheduler together — uat_test and local_test are the worked examples.
ai-worker-scheduling
hoang-phan
Applies ai-worker's scheduler discipline. Use for any task touching app/jobs/workflow_scheduler_job.rb, app/services/scheduling/, config/sidekiq.yml, config/schedule.yml, or the scheduler concurrency specs (spec/jobs/workflow_scheduler_job_spec.rb, spec/services/scheduling/).
ai-worker-review
hoang-phan
Pre-PR local review for ai-worker, covering subprocess safety, stage/enum invariants, and audit-log completeness. Use before opening a PR against this repo.
ai-worker-orchestration
hoang-phan
Applies ai-worker's job/service discipline. Use for any task touching app/jobs/, app/services/, config/sidekiq.yml, or config/schedule.yml — the scheduler and sync jobs, StageExecutor, and the Git/Github/AiCli/Prompts/Jira/Slack/Bugsnag/Brainstorm service objects.
ai-worker-detached-agent-runs
hoang-phan
Checklist for adding or changing an agent-backed fetch/sync/triage service and its job — Jira::AgentFetcher, Slack::MentionFetcher, Bugsnag::Triager on AiCli::DetachedRun. Use when a task adds a new detached-agent service, changes its result file, prompt, tool allowlist, blockers, import, sync job or schedule.yml entry, or touches AiCli::DetachedRun.
cpp-coding-standards
SeuPut0705
C++ coding standards based on the C++ Core Guidelines (isocpp.github.io). Use when writing, reviewing, or refactoring C++ code to enforce modern, safe, and idiomatic practices.
cost-aware-llm-pipeline
SeuPut0705
Cost optimization patterns for LLM API usage — model routing by task complexity, budget tracking, retry logic, and prompt caching. Use when LLM spend needs to come down, or when routing tasks across model tiers and budgets.
contract-first
SeuPut0705
Coordinate frontend/backend or service-to-service work through one authoritative machine-checkable contract (OpenAPI, AsyncAPI, Protocol Buffers, or JSON Schema), with generated consumer types and contract-verified integration. Use when parallel consumer and provider work must evolve an API or event schema without field drift, mock/production shape mismatch, or one side silently redefining the interface.
context-and-decisions
SeuPut0705
Govern stable domain context with a narrowly scoped CONTEXT.md when needed, and route hard-to-reverse decisions to ECC's existing architecture-decision-records skill. Use when domain terms require repeated explanation, one term has multiple meanings, unclear language causes module-boundary mistakes, or domain semantics must remain separate from architecture decisions.
content-hash-cache-pattern
SeuPut0705
Cache expensive file processing results using SHA-256 content hashes — path-independent, auto-invalidating, with service layer separation. Use when repeated file processing is slow and results should be cached and invalidated by content rather than path.
compose-multiplatform-patterns
SeuPut0705
Compose Multiplatform and Jetpack Compose patterns for KMP projects — state management, navigation, theming, performance, and platform-specific UI. Use when building Compose or Jetpack Compose UI, state, navigation, or theming in a KMP project.
coding-standards
SeuPut0705
Baseline cross-project coding conventions for naming, readability, immutability, and code-quality review. Use detailed frontend or backend skills for framework-specific patterns. Use when reviewing code quality or naming with no framework-specific skill that applies.
code-tour
SeuPut0705
Create CodeTour .tour files — persona-targeted, step-by-step walkthroughs with real file and line anchors. Use for onboarding tours, architecture walkthroughs, PR tours, RCA tours, and structured "explain how this works" requests. Use when the user asks for a code tour, onboarding walkthrough, PR tour, or an explanation of how a subsystem works.
clickhouse-io
SeuPut0705
ClickHouse database patterns, query optimization, analytics, and data engineering best practices for high-performance analytical workloads. Use when writing ClickHouse schemas or queries, or when an analytical query is too slow.