Dispatch a subagent to adversarially audit a PR (or the current diff) for risks, regressions, assumptions, gaps, bugs, issues, behaviour changes, leaks, and second-order consequences. Use before merging.
原文の言語: 英語
メニュー
SkillsMP は innovation-upstream/devrc から 37 件の skill を収集しています。skill を開くとソースと詳細を確認できます。
収集済み skill 37 件中 37 件を表示しています。
Dispatch a subagent to adversarially audit a PR (or the current diff) for risks, regressions, assumptions, gaps, bugs, issues, behaviour changes, leaks, and second-order consequences. Use before merging.
原文の言語: 英語
Write a canonical session-handoff doc and a copy-paste kickoff message so work resumes cleanly in a new session. Use at end of session, before a context reset, or when told to write the handoff.
原文の言語: 英語
Record what a session touched in the analyze-service index store, and write or validate an entry. The protocol /handoff follows at end of session; rarely run directly.
原文の言語: 英語
Report which shipped productivity tools are actually USED and YIELDING versus idle or DEAD. Use for "is anyone using X", "which tools are dead", "did the tool catch anything", or whether a shipped tool stuck.
原文の言語: 英語
Operate auditloop — the hosted UX-audit crawler at auditloop.zacx.dev (multi-viewport crawl, axe a11y, perf/web-vitals, run diffs and trends, vision-LLM notes, login recipes). Deploy it, manage targets/runs/plugin tokens, read the cost dashboard. Use for:…
原文の言語: 英語
Operate the homelab Tekton CI/CD platform — Pipelines/Triggers/Dashboard, the GitHub webhook (tekton-webhook.zacx.dev), the naida-ux-audit pipeline, and clawgate-ci (the real pre-merge gate). Use for: Tekton, the homelab CI platform, clawgate-ci, the Tekton…
原文の言語: 英語
Run the re-runnable naida + vetr UX-audit harnesses — walk the app's funnel, screenshot every view, capture console/network/axe findings, then re-run to verify. Local-only. Use for: UX-audit or QA-sweep naida or vetr, "make ux-audit", the funnel review, the…
原文の言語: 英語
Deterministic post-agent verification gate: re-run the AUTHORITATIVE build/typecheck/test/vet gate + git-completeness + stale-deps checks against a repo/worktree, and READ that verdict instead of trusting an agent's 'done' claim.
原文の言語: 英語
Re-enter work from the latest handoff doc: read it, re-verify it against live state, and propose ranked next steps. Use when starting a session, returning after a few days, or told to pick up where we left off.
原文の言語: 英語
Run the on-demand cross-repo initiative scan — every initiative with its momentum (active/slowing/stalled), last-touched, commits/PRs, next-step and the live tmux session hosting it; also snapshots/restores the tmux workspace across a reboot. Use for "what am…
原文の言語: 英語
Dispatch a task to opencode — the separate headless CLI agent — instead of burning this session's context on it. Preflights the brief, then launches detached. Use for "dispatch opencode", "call opencode", "run this with opencode", "hand it to flash/mimo/pro",…
原文の言語: 英語
Drive the user's LIVE, logged-in Brave browser — read the active tab's HTML, run JS in it, list/navigate tabs, screenshot the visible tab. Use when asked to look at / read / scrape / interact with a page THEY have open, act on a site they are logged into,…
原文の言語: 英語
Verify if ClickUp tasks were fully addressed by searching session transcripts for completion signals. Use when checking if recent work on assigned tasks is complete, or when asked to verify task completion status.
原文の言語: 英語
Find a past Claude Code session by keyword — searches all transcripts and returns ranked sessions with project, date, branch, the opening message, matching snippets, and the resume command. Use to recover 'the session where we did X'.
原文の言語: 英語
Operate clawgate — the self-hosted Claude Code approval UI (plus Tasks/agents/runbooks). Status, send-a-test, push/SSE logs, build and deploy a version, toggle the approval hook, manage credentials/QR. Use for: clawgate, clawgate.zacx.dev, remote approval,…
原文の言語: 英語
Read and act on ClickUp tickets and docs — comments and threads, status / assignee / due date / priority, workspace search, create tasks and subtasks, doc pages, attachments. Use for: a pasted app.clickup.com/t/ link or a bare ClickUp id, my ClickUp tickets,…
原文の言語: 英語
Operate the media download router — it files downloads by PAGE CONTEXT, not the filename. Use for: the download router, dl-route, downloads landing in the wrong folder, auto-filing downloads, the download picker/toast, a wrong route or a new site rule,…
原文の言語: 英語
Send 1:1 personalized email AS zach@vetr.com via Workspace SMTP. Use for vet-recruitment outreach to the Vetr waitlist and other genuinely personal founder email. NOT a bulk blaster — bulk owner campaigns belong on Omnisend.
原文の言語: 英語
Operate the personal activity-telemetry pipeline: 9 sources -> collector -> homelab ClickHouse activity.events -> Grafana + a per-source deadman. Query it, revive a stalled or DEAD source, deploy, validate. Use for: activity tracking, the keylogger, "where my…
原文の言語: 英語
Fix or extend the i3status-rust status bar and its status-count poller — add/reorder blocks, tune the count pills and their thresholds, debug the bar-status-poll timer and its dunst toasts. Use for: the status/i3 bar, a bar block/icon/threshold, the…
原文の言語: 英語
Decide what to automate next and SHIP it as a closed loop — score candidates, enforce the bounded-work + cheap-verifier gate, validate over real runs. Use when deciding what to automate next, when an agentic effort feels like "more harness but no value", or…
原文の言語: 英語
Operate the DURABLE initiatives board — the Postgres store, the sync, LLM recaps, the workbench viewer, the signal->initiative router and the /api/ask assistant. Use for: the initiatives viewer/board, initiatives.current/latest, the initiative store/sync,…
原文の言語: 英語
Query Prometheus/Loki/Pyroscope on a named cluster in ONE deterministic call (port-forward -> query -> teardown), with a LOUD silent-zero guard. Use for metrics/logs/profiles during an incident or perf dig, the 5xx / error-rate / latency / CPU-saturation…
原文の言語: 英語
Audit and prune the /analyze-service index store so recall keeps surfacing the RIGHT entry — evicts RESOLVED bullets, keeps every OPEN one, finds refs that resolve to two entries. Use for: prune/shrink/audit the analyze-service index, the subsystem index…
原文の言語: 英語
Audit and aggressively shrink a bloated SKILL.md body so it stops displacing the task it was loaded for — demotes detail to reference/, evicts dated history, re-measures. Use when asked to prune/shrink/trim a skill, when a skill body is huge, or when its…
原文の言語: 英語
READ the weekly repo chief-of-staff proposals and evaluate them WITH Zach; also re-run the scan, tune the signals/model/repo-list, manage the timer. Use for: repo proposals, the repo chief-of-staff, "the ideas the agent emailed me", triaging the weekly repo…
原文の言語: 英語
Live cross-host view of every tmux window on workbench + laptop — which have Claude Code running, what each is doing, which ones are WAITING ON A HUMAN (asked a question / blocked on a modal / out of context), how stale it is, plus the clawgate approval…
原文の言語: 英語
Query and operate the self-hosted Signal chat pipeline (phone -> signal-cli-rest-api -> homelab Postgres + MinIO attachments), and DRAFT outbound replies for clawgate approval. Use for: my Signal messages, signal chat, who messaged me on Signal, search my…
原文の言語: 英語
Point at ONE tmux window by codename, hotkey or address (Gold, Alt+Shift+G, scratch2:@81), and rank the windows STRANDED past a threshold, longest-waited first. Use for: which window is Gold, what does Alt+p open, which window has been stranded / unanswered…
原文の言語: 英語
Recon a service/subsystem fast: locate where it lives, load its config, check its live state and recent changes — then optionally do the follow-on task. Replaces hand-typed 'analyze the {redis,minio,flux,bastion,monitoring,…} setup, then …' recon.
原文の言語: 英語
Re-run the espanso snippet usage audit — cross real keylog fire counts against transcript demand for a per-snippet keep/retune/prune VERDICT, then tune the snippets in nix/home.nix.
原文の言語: 英語
One-shot fleet-wide status sweep across every active repo and cluster — your approved-mergeable/red/conflicting PRs, in-flight deploys and rollouts, and firing alerts split per cluster, as an action-first roll-up. Use for "standup", "fleet status", "what's in…
原文の言語: 英語
Audit and fix the devrc desktop/terminal setup — home-manager, NixOS, i3, tmux + fuzzyclaw, zsh/direnv, gruvbox theming — reporting cross-cutting drift then fixing it. Use for: audit or improve my shell/tmux/i3/desktop config, a keybinding or theme…
原文の言語: 英語
Drive the i3 window manager — query windows/workspaces/layouts, focus/move/resize/arrange, screenshot the desktop and analyze it, type/click/scroll and read the clipboard via xdotool. Use for: what's on my screen, which workspace, focus or move a window,…
原文の言語: 英語
Query and operate the self-hosted mail inbox (Gmail -> inbox.zacx.dev -> homelab Postgres `mail`) and its mail-actions automation, and SEND email as the operator via Gmail SMTP. Use for: my mail inbox, inbox.zacx.dev, forwarded email, the mail-receiver, the…
原文の言語: 英語
Audit and aggressively shrink a project's auto-memory index (MEMORY.md) — it costs tokens EVERY session and silently drops content past its cap. Runs scripts/memory-audit.py, classifies every bullet, rewrites atomically, re-measures. Use when MEMORY.md is…
原文の言語: 英語
Manage the SGLang inference server on the workbench cluster (ns promptver): swap the served model, check pod/model/GPU health, port-forward, scale up/down, read logs, diagnose OOM or GPU contention with ComfyUI. Use for: SGLang, sglang-server, changing the…
原文の言語: 英語