dontbesilent2025/dbskill
24 skills132k installs152k starsGitHub
Install
npx skills add https://github.com/dontbesilent2025/dbskillSkills in this repo
1DbsDbs is the dontbesilent business toolbox front door for solo creators and indie operators who use slash commands and natural language to get structured help. The skill’s only job is to read intent signals—benchmark hunting, hook writing, AI-voice checks, procrastination, fuzzy goals, decision journaling, or tri-platform Agent desk cleanup—and hand off to the matching specialized skill in the same repo family. That makes it journey-wide: the same router works when you are scoping an offer in Validate, drafting growth content in Grow, or untangling Claude Code versus Codex setup in Build. It deliberately avoids giving business answers itself, which prevents blended advice and keeps each downstream diagnostic intact. Prism lists it under Validate and scope as the shelf where ambiguous “what should I do?” questions land first, even though many routes immediately serve Launch and Grow workflows.9.2kinstalls2Dbs Contentdbs-content is a dontbesilent agent skill for solo knowledge creators who already picked a topic and need a ruthless editor-coach, not a draft machine. It enforces that the human writes while the agent judges whether the angle, medium (short video, graphic carousel, long video, live, article), and voice fit the idea—grounded in five philosophies: text discipline beats AI stigma, titles as cognitive hooks, effort times understanding beats volume spirals, monetizable product before publishing, and clarity requires understanding first. Phase 1 intake confirms topic and format; Phase 2 matches topic traits to recommended formats and platforms; Phase 3 reviews copy when a draft exists. Triggers include /dbs-content and natural phrases like “how should I make this content.” Best after validation when you have something to sell; skip if you only want full ghostwritten posts.8.9kinstalls3Dbs Diagnosisdbs-diagnosis is a journey-wide agent skill grounded in dontbesilent’s commercial ontology: it treats the business model as an objective system and your job as aligning inputs to that machine—not chasing heroic hustle advice. On start it forces a mode choice between consultation (you bring one sharp question) and checkup (no specific question, full model scan). Consultation begins by letting you speak uninterrupted, then classifies whether you need factual how-to or deeper reframing; the skill’s default stance is that most “business problems” are mis-posed and should be dissolved before solutioning. The embedded philosophy covers pricing ladders (e.g., 5–15× gap between lead and profit offers), skepticism of vanity traffic, and the claim that execution plus model beats raw IQ for cash flow. Solo builders use it when stuck on monetization, offer design, or whether to pivot—especially if you keep researching tactics instead of fixing structure. Outputs are reframed questions, model critiques, or a structured diagnosis report depending on mode—not code.8.8kinstalls4Dbs Benchmarkdbs-benchmark is a dontbesilent-style benchmark analysis skill for solo builders in the 0→1 phase who need a ruthless filter for who to imitate—not another personality quiz about “what fits you.” The agent runs a structured flow: clarify current or intended business, then pass candidates through five filters covering profitability (roughly ten times your current or expected income), whether you can see the full money chain, whether you can realistically replicate acquisition and delivery, deliberate exclusion of self-focused excuses, and a single north star of high profit rather than vanity metrics like fans or hype. The philosophy treats imitation as the correct early-stage strategy and treats differentiation as a later luxury. It is delivered in Chinese with English trigger phrases, ideal for indie operators choosing SaaS, content, or e-commerce plays. Use it before heavy build spend when you are tempted to “just be yourself” instead of studying a proven counterpart.8.8kinstalls5Dbs Deconstructdbs-deconstruct is a dontbesilent methodology skill that forces vague commercial language into testable, atomic pieces. Solo and indie builders in Chinese-language markets invoke it when a pitch deck, guru thread, or their own notes use terms like private traffic, moats, or knowledge monetization without operational clarity. The agent walks a phased ritual: capture the term or confusing sentence, analyze how different speakers use it, then rebuild meaning through seven ontology tables inspired by Wittgenstein and Austrian economics. It distinguishes questions with textual answers from problems that only resolve through practice—so you stop treating “how to make money” like a Googleable FAQ. The skill is prose-heavy and editorial, not an API integration; it pairs well with validation and positioning work once definitions are honest. Use it whenever language is steering your strategy faster than evidence.8.8kinstalls6Dbs Hookdbs-hook is a dontbesilent agent skill for creators shipping short-form video who need openings that earn watch-through in the first seconds. It treats the hook as a standalone trial of the video, not an extension of the title, and follows a fixed philosophy: combine topic, attention hook, and credibility, prefer suspense over premature answers, and write lines that sound natural when spoken aloud. The agent first collects the full script, then runs content-quality gates—rejecting empty bodies, thin copy, or valueless rants—before suggesting concrete opening rewrites with contrasting bad versus good examples. Solo builders growing on Douyin-style platforms use it when a draft feels flat at second one even though the middle might be fine.8.1kinstalls7Dbs Actiondbs-action is a dontbesilent execution-diagnosis agent skill grounded in Adlerian individual psychology and founder-focused observational rules. It exists for solo builders who can describe the right next step yet chronically avoid it—not because they lack tactics, but because avoidance serves a hidden purpose such as protecting identity or sidestepping failure. The skill runs a structured conversation: let the user state the concrete stall, scan for named signals like treating people and tools as “simulators” instead of executing, then deliver a direct diagnosis rather than encouragement. Content is primarily authored in Chinese with English trigger phrases, which matters for audience fit on Prism. It applies across the journey whenever execution collapses—stalling validation tasks, avoiding launch distribution, or iterating in Operate—while listing Validate as the primary shelf because scope commitment is where the gap is most visible. It pairs well with planning skills only after you accept the emotional/economic reason you postponed action.8.1kinstalls8Dbs Xhs Titledbs-xhs-title is an agent skill for solo builders and creators who publish on Xiaohongshu (RED). Instead of free-form headline chat, it walks you through clarifying topic and audience, then selects five to eight formulas from a library of seventy-five patterns grouped into twelve trigger types—cognitive conflict, curiosity gaps, loss aversion, identity hooks, numbers, result promises, social proof, controversy, scenarios, CTAs, authority, and participation prompts. Each suggested title must map to a numbered formula so you learn which psychological lever fits your niche, whether you are sharing opinions, tutorials,避坑 warnings, or vertical母婴/餐饮 content. Triggers include /dbs-xhs-title, 小红书标题, and English RED title requests. Best when you already know what you want to say but need structured, platform-native packaging that has been validated on RED, not one-off clever lines.6.7kinstalls9Dbs Ai Checkdbs-ai-check is a diagnostic agent skill for founders and marketers who publish Chinese (and mixed) copy and want to see where text feels machine-perfect rather than human. It walks the document in reading order, quotes each suspect passage, labels severity, and summarizes the worst patterns without automatically rewriting. The philosophy is explicit: detecting AI texture is about clarity of voice, not disguising authorship. Triggers include /dbs-ai-check, requests to check for AI flavor, or English phrases like “does this sound like AI.” When the user opts into edits, the skill guides with questions per finding instead of bulk paraphrase. Use it in Grow when newsletters, landing copy, or social posts are ready for a sanity pass, and in Launch when distribution copy must sound intentional.6.3kinstalls10Dbs Slowisfastdbs-slowisfast is a journey-wide coaching skill from the dontbesilent “slow is fast” philosophy for entrepreneurs who suspect they are optimizing for speed at the cost of durable advantage. It does not romanticize slowness; it helps you separate work that should stay fast from work where deliberate friction builds judgment, reusable assets, and easier next iterations. The agent walks through your concrete plan or habit, audits bypassed friction, and tests whether the approach leaves behind systems, libraries, or understanding that compound. That maps cleanly to solo builders deciding between batch AI content, one-off hacks, and manual passes that create templates, benchmarks, and customer insight. Use it whenever a method feels efficient but empty, or when you ask if you are moving too fast before committing budget, brand, or architecture.6.2kinstalls11Dbs Chatroomdbs-chatroom is a journey-wide facilitation skill for solo builders who want structured debate instead of a single generic assistant answer. Trigger with /dbs-chatroom, /定向聊天室, or the Chinese phrase 定向聊天室. The host either accepts named experts or recommends three to five thinkers, then waits for explicit confirmation—a deliberate gate so automation does not skip human intent. Confirmed runs spawn one Agent per expert in parallel using a shared honesty and style template, then a judge pass from the main model critiques insight, blind spots, and deliverable actions. It fits early Idea research, Validate scope debates, and Grow content strategy when you need opposing mental models before you commit engineering time. Chinese-first UX; agents should preserve trigger phrases and stop points exactly as documented.6.1kinstalls12Dbs Chatroom Austriandbs-chatroom-austrian is an agent skill that hosts an Austrian-school economics chatroom. A solo builder or curious operator poses a topic; the host launches Hayek and Mises as parallel sub-agents, then Claude synthesizes as judge with concrete takeaways. It fits early journey research when you want principled pushback on regulation, markets, pricing, or coordination design—not implementation checklists. Triggers include slash commands and Chinese phrases for 奥派聊天室. The methodology encodes honest disagreement rules (Hayek may concede limits of spontaneous order; Mises stays axiomatic). Outputs are readable multi-voice transcripts plus actionable judgment, making it useful for sharpening positioning, pricing narratives, or policy-heavy indie bets before validation spend.6kinstalls13Dbs Agent Migrationdbs-agent-migration is dontbesilent’s agent workbench migration skill for builders whose Claude Code, Codex, and Grok setups have drifted into half-migrated chaos. It is not an installer and not a script runner—it is a procedural migration with audit, consolidation, naming normalization, bridge generation, and structural verification. The skill addresses structural migration: Claude to Codex, Codex back to Claude, Grok in either direction, or all three hosts aligned from a tangled starting point. It will not run commercial diagnosis, rewrite your knowledge base, judge one skill’s methodology, or write marketing copy. Route here when users ask to migrate to Codex or Grok, fix AGENTS.md, unify Claude and Codex and Grok, or clean scattered local skills. The outcome is an maintainable agent workspace where which file is authoritative is obvious and bridges keep hosts consistent. Solo builders wearing every hat use this when agent config debt slows every session—before adding more skills on top of a broken layout.5.9kinstalls14Dbs Reportdbs-report is a dbskill deliverable generator for solo consultants and indie builders who run structured diagnosis sessions and save them with dbs-save. Instead of screenshotting chat, the agent reads archived markdown under ~/.dbs/sessions per project slug, orders files chronologically, parses frontmatter (status, title, source_skill, next_skill), and produces one indexed markdown report suitable for partners or future review. Triggers include /dbs-report, packaging requests, and phrases like delivering something for a co-founder. The skill enforces trust: the report is a merge of user-confirmed saves, not a fresh AI narrative. It pairs naturally after repeated saves when you need versioned, date-stamped documentation from an ongoing diagnostic workflow.4.8kinstalls15Dbs Savedbs-save is a journey-wide agent skill in the dbskill stack that records—not performs—business diagnosis. Solo builders running Chinese-language diagnostic workflows with dbs-diagnosis and related skills use it when a session surfaces conclusions, vetoed directions, and recommended next steps that must survive until the next chat. The skill extracts a title, packs state into markdown with explicit metadata, and stores it under a project slug derived from the working directory or user override. That turns one-off Q&A into an accumulating record so you do not re-explain customer context every week. Invoke via /dbs-save, /存档, or phrases like save this diagnosis. It explicitly defers diagnosis to other skills and only runs when there is substance to persist. Ideal for indie operators juggling multiple small ventures who want local, user-controlled memory without a cloud CRM.4.8kinstalls16Dbs Restoredbs-restore is a continuity skill in the dontbesilent dbskill family that pulls the most recent locally saved diagnosis snapshot so a solo builder can pick up exactly where the last session stopped. It answers slash commands and natural phrases like continuing from last time or asking where diagnosis left off, then renders status, title, source skill, and suggested next skill from the archive metadata. The skill deliberately avoids starting new analysis or silently chaining into unrelated skills; it only restores memory. Snapshots are keyed by project slug derived from the working directory basename, with deterministic ordering by YYYYMMDD-HHMMSS filename prefixes instead of unreliable file mtimes. When no archive exists, it guides you toward other slugs with recent activity. Use it whenever long-running business or product diagnoses span days and you rely on agents that otherwise forget prior conclusions.4.8kinstalls17Dbs Goaldbs-goal is a journey-wide agent skill from the dontbesilent stack that stops goal language from spinning without doing work. Solo builders often arrive with polished-sounding aims that cannot pick a next task or an finish line; the skill treats that as “engine idling” and runs a structured audit inspired by Wittgenstein’s idea that meaning shows up in use. You first paste the goal exactly as you think it—no smoothing—then the agent walks through usability tests for concreteness, falsifiability, and completion, stripping hollow words like “systematic”, “influence”, or “long-term” when they add no constraint. Good targets share at least three of five family traits: pointable deliverables, possible failure, a done state, grammatical clarity (who does what to what standard), and fit with your real constraints. Use it in Idea when exploring direction, in Validate before locking scope, in Build when a PM spec still reads like a wish, and in Grow when positioning drifts. It outputs checkable deliverables and completion cues your agent or you can execute—not motivational copy.4.5kinstalls18Dbs Learningdbs-learning is a dontbesilent agent skill for solo builders who want structured study without a fixed syllabus. You name a topic; the agent splits it into a continuous chain of markdown lessons and maintains an adaptive learning gradient—each installment responds to how you reacted to the last one (depth, angle, pacing). It is not a substitute for business diagnosis or ghostwritten content; those cases are routed to sibling dbskill capabilities. Files land under 学习课题/{课题名}/ in an active repo when the cwd looks like a project, otherwise under ~/Documents/dbskill-learning with a clear notice when the current folder is too generic. Invoke when you want a curriculum that evolves with you rather than a single explainer or when you want dbskill methodology broken into teachable steps.3.4kinstalls19Dbs Good Questiondbs-good-question is a journey-wide facilitation skill from dontbesilent that reframes messy questions—why sales stall, whether a workflow can be automated—into problem specifications agents can criticize and verify. Instead of answering oversized whys immediately, the agent nails an observable phenomenon, surfaces conflicting signals, and documents object, goal, variables, constraints, and falsifiable feedback. It then rates how far automation can go without pretending closed-loop fixes exist when feedback only lives in human conversations. Solo builders benefit before Idea validation, during Grow diagnostics on content or funnel metrics, and whenever they hand a task to Claude Code without a crisp brief. The philosophy rejects premature certainty: missing data yields minimal follow-up questions or smallest observational next steps. Use slash commands or natural-language triggers when you feel stuck in analysis paralysis or before spawning subagents on underspecified work.3.2kinstalls20Dbs Decisiondbs-decision is dontbesilent’s personal decision-system agent skill. It turns any domain you care about for months or years—business, relationships, health, career, learning, or investing—into a local Markdown knowledge project with a strict four-layer layout, source tagging, append-only facts, slowly evolving pattern notes, frozen snapshots, and a living queue of open questions and hypotheses. The agent’s job is knowledge engineering: write what happened into the right layer, keep chat-transient detail out of durable judgment, and make later sessions and future-you able to pick up the thread from files instead of memory. Triggers include /dbs-decision, /决策系统, /决策立案, /结果回填, and /状态画像. It fits solo and indie builders who already work in Claude Code, Cursor, or similar agents and want procedural discipline without outsourcing the actual call. Use it whenever a decision thread will outlive one conversation—not for one-off prompts with no archive.2.5kinstalls21Chatroom Austrianchatroom-austrian is an agent skill that hosts a multi-role Austrian economics chatroom for solo and indie builders who want disciplined debate before they commit to strategy. You pose a topic—regulation, pricing, platform design, or macro narrative—and the skill coordinates Friedrich Hayek and Ludwig von Mises as separate sub-agents while Claude moderates, flags blind spots, and closes with concrete judgments you can act on. The workflow is explicit: welcome, fire both economist agents in parallel, render their replies, then deliver a quality-focused synthesis that resists lazy theory-matching. It fits early journey work when you are still mapping incentives and information flows, and it remains useful in Validate and Grow when you revisit unit economics or distribution assumptions. The skill is procedural knowledge packaged as SKILL.md orchestration, not an MCP server or data feed; expect to bring your own question and tolerate normative economic framing.1.9kinstalls22Dbskill Upgradedbskill-upgrade is a meta maintenance skill for solo builders who rely on the dbskill bundle inside Claude-style skill directories. When you invoke /dbskill-upgrade or the Chinese upgrade triggers, the agent follows a scripted bash workflow: confirm dbskill lives under ~/.claude/skills, read the current VERSION file, pull the remote VERSION from GitHub raw, and exit early if you are already current. If an upgrade is needed, it copies existing dbs* folders into a dated backup, shallow-clones dontbesilent2025/dbskill into a temp directory, and replaces the old install with the freshly downloaded tree. The skill emphasizes transparent version reporting so you can see what changed between OLD_VERSION and REMOTE_VERSION before trusting new behavior. It assumes network access for curl and git, filesystem writes under the home skills path, and comfort restoring from backup if a clone fails. This is not for designing product features—it keeps your agent skill substrate aligned with upstream fixes and additions.1.7kinstalls23Dbs Content Systemdbs-content-system is an acceptance and workflow skill for solo builders who turn messy source manuscripts, tweets, and numbered shorts into a governed content factory. It tells your agent when the system is “usable” without demanding full corpus completion on day one: scaffold the repo, copy raw assets into the intake zone, generate source candidates and pending queues, produce the first content units and theme map, and wire relationship and deduplication indexes. The skill is strict about placeholder rejection—questions cannot echo titles, claims cannot be keyword lists, and case/solution fields must carry real steps. It encodes operational scripts for link maps and duplicate candidates and expects Obsidian visibility after tooling runs. Use it when you are building a long-horizon repurposing stack and need hard gates before batch automation.1.4kinstalls24Dbs Unblockdbs-unblock is a journey-wide agent skill from dontbesilent that diagnoses execution blocks for solo founders using an Adlerian psychology frame blended with blunt entrepreneurial observation. Install it when you already know the next business step—publish, sell, ship, outreach—but keep researching, simulating Claude Code runs, or asking for one more opinion instead of acting. The skill runs a phased ritual: elicit the specific stall, listen without premature fixes, then match language to diagnostic signals such as the execution simulator pattern where preparation substitutes for contact with failure. It treats most apparent startup problems as disguised identity and risk-avoidance problems, emphasizes separating feelings from action, and refuses empty encouragement. It is written primarily in Chinese with English trigger phrases in the frontmatter, so it fits indie builders working in that voice who want an agent mirror, not a hype coach. Use it at any journey phase whenever resistance shows up before commit, during launch dread, or when growth tasks get infinitely deferred.631installs