Claude-AI-Tools-Ventura-County
- 19 skills
- 0 followers
- 11 hours ago last updated
- ▌ Giantbrains · claude-ai-tools-ventura-countyRouter for the Giant Brains suite: stress-test one doc (a plan, spec, proposal, or migration doc) by running the two or three suite lenses that match the doc's stage, report-only, then synthesizing one combined verdict. Triages in a single message (which doc, what stage), routes by stage — draft: take-a-step-back, iron-triangle, blast-radius; in progress: blast-radius, iron-triangle; complete: decision-record audit, bottom-line — and never edits the doc: writers (record-decision, linear; phase-qa when giant-brains-swe-skills is installed) are offered afterward as explicit opt-ins. Trigger when the user invokes /giantbrains, says "stress test this plan/doc", "pressure-test this", "run the battery", "run all the lenses/brains against this", or asks for a full multi-angle review of a planning doc. Do NOT trigger for a single quick decision (route to the one matching skill), line-by-line code or correctness review (/code-review), or a request to edit, reformat, or rewrite the doc itself.
- ▌ Linear · claude-ai-tools-ventura-countyRestructure multi-step output into a single linear, numbered, step-by-step plan with branches as indented sub-bullets — concise but detailed. Use this whenever a response (yours or another agent's) contains multiple steps, instructions, or actions scattered across verbose prose, split across sections, or buried in explanations. Trigger when the user says "linear", "linearize", "make this step-by-step", "give me a plan", "what do I actually do", "put this in order", or pastes a verbose response and asks for the actionable sequence — including completion/status messages whose remaining work is scattered across "what I didn't do", "open items", and "next steps" sections. Also self-trigger when about to give 3+ procedural actions the user must execute themselves — but not for option comparisons, conceptual explanations, or summaries with no remaining work to execute.
- ▌ Btw · claude-ai-tools-ventura-county bundleStarts and maintains a manual session attention firewall for Claude Code. Invoke only with /btw when the user explicitly wants focus mode, no side quests, or a maximum of three active tasks. Create one append-only Markdown log per BTW focus session, capture the opening user prompt plus repository and branch context, and park assistant-initiated off-task findings as Action, Review Later, or Interesting instead of surfacing them in assistant prose. After activation, treat plain-language "btw check" as an in-session re-anchor rather than a new invocation.
- ▌ Feynman · claude-ai-tools-ventura-countyConvert a highly technical document or subject into a plain, accurate, layman-friendly explanation using Feynman-style teaching moves — everyday analogy, a first-principles ladder, jargon stripping, and honesty about the limits. Use this whenever the user wants to "explain this simply", "make this understandable to a non-technical audience", "translate this for stakeholders / execs / new hires / customers", "ELI5 this doc", "de-jargon this", give me the "plain-English version", "what does this actually mean", or hands over a dense technical doc, spec, paper, RFC, or codebase writeup and wants it made digestible — even if they don't say the word "Feynman". Produces a layered explanation (one-sentence core, then gist, then full plain walkthrough), a jargon-to-plain glossary bridge, analogies that each name where they break down, and a flagged list of anything the source left genuinely unclear. Do NOT use this to merely shorten a document that is already clear (that is a summary, not a translation), to write mar
- ▌ Myriad · claude-ai-tools-ventura-county bundleEnd-of-day triage funnel for messy agent output. Takes a wall-of-text agent completion message that blends finished work, blockers, gated next steps, and stray follow-ups, and sorts it into four buckets: Completed (shipped), Critical Blockers (what stops you now), Awaiting Your Call (gated next steps and decision prompts that need your go/no-go), and the "myriad" (nice-to-haves, ideas, dupes, someday-maybes). Shows a clean Done / Broken / Awaiting summary, nudges you with current git status so uncommitted work never gets abandoned, and parks the long tail in a durable weekly backlog file using an idempotent, verify-after-write append so nothing is ever lost or double-logged. Trigger when handed a long, mixed agent response at the end of a work session, or on /myriad.
- ▌ Bottom Line · claude-ai-tools-ventura-countyDistill long, verbose, or repetitive output—especially from Claude or other AI agents comparing options, tradeoffs, and recommendations—into a short, deduplicated bottom-line summary. Surface the real decision, a clear recommendation, and any choice that would be hard to reverse later. Trigger when content is decision- or tradeoff-oriented and the user asks for the bottom line, TLDR, gist, takeaway, to cut to the chase, summarize cleanly, says it is too long/verbose/repetitive, or asks what they actually need to decide or might regret. Also self-trigger when prior agent output is long-winded, hedged, or repetitive and a decision needs to be pulled out clearly, even without those exact words. Do not trigger when the user wants procedural steps, commands, or an execution sequence; that is /linear's job, not this skill's.
- ▌ Worth It · claude-ai-tools-ventura-countyJudge whether a planned feature or refactor is worth building — whether the payoff justifies the cost and actually moves a needle that matters. Judges benefit against cost across the constituencies a change can serve (end-user UX, human operator experience, codebase maintainability/durability, contributor velocity, business/strategic obligation), discounts benefit that is asserted but unmeasured, nets a gain on one front against the cost it adds on another, and weighs the whole thing against the do-nothing baseline, the timing window, and the next-best use of the same effort. Trigger when the user asks "is this worth doing/building," "is it worth it," "is this worthwhile," "does this move the needle," "what's the payoff/ROI," "should we even bother," is weighing whether a feature or refactor earns its place, or hands over a plan doc whose open question is its *value* rather than its engineering or its risk. Assumes the frame is already right (that is take-a-step-back); it leans on blast-radius and iron-triang
- ▌ Loose Ends · claude-ai-tools-ventura-countyPost-work completeness sweep and execution (alias: close-loop). Check what was forgotten before declaring work done, and actively execute the final steps. Use only after work exists: a diff, draft, or completed change session. Compare what was delivered against what was requested, then look for dropped requirements, unrun verification, and leftover scaffolding. Actively offer to fix these gaps, run linters, sync docs, and stage/commit/push the work to git. Trigger when the user asks "close loop", "close-loop", “what did I forget,” “did I miss anything,” “is this done,” “ready to ship/PR?”, or is about to call a multi-file or multi-requirement task complete. Also self-trigger before reporting substantial multi-step work finished. Do not trigger before work exists; if the question is about a plan, decision, or approach, route to take-a-step-back. Do not use for phased-plan QA (phase-qa) or line-by-line code review (/code-review). This skill hunts what is absent, not what is wrong.
- ▌ Snapshot · claude-ai-tools-ventura-countySave the most recent substantive response — plus session metadata, recent test findings, and phase status — to a persistent, additive snapshot.md file so the user can re-find and resume the session later (next morning, after a crash, or across chat tabs). Trigger whenever the user says "snapshot", "snapshot this", "save this session", "save our progress", "save before I go to bed / sign off / wrap up", "checkpoint this", or expresses worry about losing work, losing the chat, or VS Code crashing. Also offer it proactively when a long working session reaches a natural stopping point with unsaved findings. Use only to preserve session progress — not when the user wants to save or export a specific artifact (a file, PDF, image, code snippet, or document).
- ▌ Rabbit Hole · claude-ai-tools-ventura-county bundleStop the drip of one-more-thing on a task that was supposed to be simple, and replace it with a single end-to-end triage that puts every issue on the table at once. Use when an agent keeps surfacing problems piecemeal (death by a thousand cuts), keeps adding unrequested work (scope creep, gold-plating, "while I'm here..."), or wanders off the asked task into refactors and tangents. Trigger when the user says "you keep finding one more thing", "stop nickel-and-diming this", "this was supposed to be simple", "stop going down rabbit holes", "just do a full pass and tell me everything", "triage it all at once", or invokes /rabbit-hole. Also self-trigger when you are about to raise a third unsolicited mid-task finding on a task the user explicitly framed as small ("quick", "simple", "one-line", "just"). Do NOT fire on genuinely open-ended or exploratory work where incremental discovery is the point, or during an active debugging session where each finding narrows the next step.
- ▌ Finish Line · claude-ai-tools-ventura-countyUse when the user asks to close, wrap up, finalize, ship, or define "done" for a plan, chapter, PR, milestone, or project; or says any of: "finish line", "definition of done", "done list", "stop adding scope", "no new items", "wrap it up", "close the chapter". Recommends the shortest safe path from the current branch to the next checkpoint: remaining work, governance compliance, and ship-safety blockers only — everything else is parked in /PARKED for later pickup, never narrated in chat.
- ▌ Stay Focused · claude-ai-tools-ventura-county bundleLock a whole working session to one anchor task so it can't be forgotten or side-tracked, and force every reply to lead with that task's live status in minimum-viable text. Type "stay focused on <task>" (or "stay on track for <task>") at the start of a session; every later reply then opens with the current state of that anchor — fixed? shipped? blocked on an unmerged branch? — followed by a horizontal rule and a tight rabbit-hole-style body (What's Shipped / What's Blocked / Recommended Next Steps / Optional Next Steps / Only Useful FYIs), each bucket one line, empty buckets dropped. Trigger when the user says "stay focused on X", "stay on track for X", "keep me focused on X", "/stay-focused", or asks you to stop wandering off the session's original task. If no anchor is named and none is set, ask what to lock onto rather than guessing. Turn it off with "/stay-focused off", "stop staying focused", or "drop the focus". Pairs with an always-on hook (see HOOKS.md) that re-injects the anchor every turn so it surv
- ▌ Blast Radius · claude-ai-tools-ventura-countyFlag the true cost of a recommendation before the user commits, so a large or hard-to-reverse change is never presented like a minor tweak. Fire automatically when you are about to recommend a refactor, a migration, a schema change, a dependency or framework swap, a public API or contract change, a move across service boundaries, or anything else expensive to adopt or painful to undo. Also trigger when the user asks whether something is "a big change," "a major refactor," "risky," "breaking," "hard to unwind," "a one-way door," "painting us into a corner," or "more work than it sounds." Skip only when the user explicitly asks for the best path regardless of cost.
- ▌ Iron Triangle · claude-ai-tools-ventura-countyForce an explicit speed-versus-cost-versus-quality tradeoff when an operator is under pressure to ship something faster, cheaper, or better than the constraints actually allow. Use this when the user is weighing "do it right" versus "just ship it," is tempted by an ambitious rework before doing the real work, is handed a fixed date without more resources, or says things like "we don't have time to do this properly," "fast, cheap, or good," "is it worth doing the right way," "should we cut scope," or "we can't afford to slow down." It names which corner of the triangle is being sacrificed, checks whether that sacrifice stays contained or compounds, and points to scope as the release valve.
- ▌ Auto Improve · claude-ai-tools-ventura-county bundleThe EXECUTOR of the optimization pair: run a bounded, self-verifying loop that measurably improves a single file, only AFTER a measurable contract already exists. Fire when a scalar metric, a strong correctness oracle, and a fixed budget are ALREADY defined — typically handed off by baseline-spec, or when the user says they have a benchmark/eval and a baseline — or when the user explicitly invokes the loop: 'run the auto-improve loop,' 'let it iterate until it's faster,' 'hill-climb this,' 'auto-tune against my benchmark,' 'Karpathy-style optimization loop.' For a cold-start request that has NOT yet named a number, gate, and stop — 'optimize this,' 'make it faster,' 'make it cleaner,' 'tune this prompt' — do NOT run this loop; defer to baseline-spec to define the contract first. Re-gate on the three preconditions before every run; if any is missing, refuse and name it. Skip entirely when the goal is subjective with no number behind it.
- ▌ Better Options · claude-ai-tools-ventura-countyPressure-test an option set before the operator picks. Catches an AI agent — this one included — conflating two separate problems into one menu, or quietly narrowing the field to a false dilemma that skips the boring option. Borrows debug-mantra's falsify-the-hypothesis rigor (disprove each option before trusting it) then ponytail's minimalism bias (of what survives, favor the smallest) to surface the option nobody named because it wasn't impressive enough. Trigger when an agent — this one included — or a user presents two or more options for a non-trivial engineering decision, especially when every option involves new abstractions, infrastructure, or a rewrite; when the framing sounds like "you have to either X or Y," "there's no way around it," or "we'd need to..."; or when the user asks "is that really the only way," "feels like you're overcomplicating this," "are these actually the same problem," or "what's the simplest version that still works." Also self-trigger internally before presenting any option m
- ▌ Baseline Spec · claude-ai-tools-ventura-countyThe DEFINER and entry point of the optimization pair: turn a vague "make it better" into a measurable contract before any optimization or eval begins — a metric, an un-gameable correctness oracle, a budget, and a captured baseline. Fire FIRST, before auto-improve, on any cold-start optimization request that has named no number, no correctness gate, and no stopping rule: "optimize this," "make it faster / cleaner / tighter / more engaging," "tune this prompt," "shrink this," "improve the copy." This skill runs before the executor; do not let the loop fire on an undefined request. Default to a one-shot spec sheet that names the metric, the oracle, the budget, and the single missing fact; drop into one-question-at-a-time interrogation only if the operator engages. Refuse to optimize until all three pillars are defined. Skip only when the metric, oracle, and budget are already specified — then hand straight to auto-improve.
- ▌ Record Decision · claude-ai-tools-ventura-countyRecord or update durable decision records in a repo — the call, the assumption it bets on, the expected signal, reversibility, and a revisit trigger — then keep the record and the project docs current as findings arrive. Use ONLY when the choice clears the threshold: a meaningful bet that could be wrong, Costly or One-way-door reversibility, or a future signal worth checking. Above that threshold, trigger RECORD mode when the user says "record this decision", "log this", "write this down", commits to an option after weighing tradeoffs ("let's go with X" / "we're going with X"), or a decision emerges from take-a-step-back, iron-triangle, blast-radius, or bottom-line output. Trigger UPDATE mode when evidence arrives, an experiment concludes, an assumption is confirmed or broken, a recorded decision is reversed or superseded, or the user says "update the decision log" / "the migration worked" / "that bet didn't pay off". Never trigger for formatting, naming, lint-level, or five-minute-revert choices.
- ▌ Take A Step Back · claude-ai-tools-ventura-countyRun a brief but meaningful decision reset before committing to a plan, recommendation, or next step. Use this when the user is about to make a choice, accept an LLM recommendation, commit to a direction, or push forward on a plan that may deserve a quick 360 re-evaluation first. Especially useful when assumptions may be hidden, the problem may be framed too narrowly, urgency may be driving the choice, or the path may be hard to reverse later.