Total 57,970 skills
Showing 12 of 57970 skills
Analyze candidate models before adapter implementation. Determine model implementation source (transformers or model-local), structural features, layer-by-layer loading requirements, and MoE fused weight risks. Use this skill when the user wants to: (1) assess model adaptation feasibility before creating msModelSlim adapters, (2) analyze model structure and type classification, (3) evaluate MoE compatibility for quantization. Trigger: user mentions "model analysis", "msModelSlim", "adapter", "transformers", "MoE", "layer-by-layer", "model assessment", "feasibility", "模型分析", "适配可行性", "模型评估", "MoE分析"
Conduct a structured business-logic interview for a new feature, asking questions iteratively until shared understanding is reached, then save a clean spec doc to `docs/<feature-slug>-business-logic.md`. The skill always writes the spec to disk and always stops there — it never continues into implementation planning. Use during planning — before any code or file paths are discussed — whenever the user wants to add a feature, change a flow, or scope new behavior. Trigger on phrases like "I want to add X", "let's plan a new feature", "we need to support Y", "I'm thinking about a feature for Z", "scope out…", "let's figure out what should happen when…", or any time the user describes a new capability and the next move is clarifying *what* it should do (not *how* to build it). Triggers even if the user doesn't say "plan mode" or "business logic" — if a feature is being scoped, this is the right skill. Refuses to discuss code, file paths, schemas, libraries, or implementation; redirects back to user behavior and business rules. Output is a saved markdown spec; running the implementation-planning skill afterward is a separate, user-initiated step.
Break an implementation plan markdown file into a sequenced, dependency-annotated, checkbox-tracked task list saved as a sibling doc (e.g. `docs/<feature>-tasks.md`). Each task is annotated with the lower-numbered tasks it depends on (or marked dependency-free) and ordered topologically, so independent tasks can be implemented in parallel. The task list is the execution contract — each task is one logical commit that bundles the code change **and** the ticked checkbox together (never a separate "tick-only" commit). Use this whenever the user has an implementation plan (file like `docs/<thing>-plan.md`, an `## Implementation plan` section, or a freshly written design doc) and wants a step-by-step todo list to drive the build. Trigger on phrases like "break this plan into tasks", "make a task list", "generate the tasks doc", "turn this plan into checkboxes", "split this into steps I can implement one by one", or whenever a plan exists and the next move is execution — even if the user doesn't explicitly say "task list".
SQLx + PostgreSQL v17 database guide. Queries, migrations, ENUMs, JSONB, transactions.
Research trends deeply across social platforms and search engines, preserving source evidence and producing a cited, reproducible report. Use for multi-platform trend discovery, content intelligence, audience research, or long-running topic monitoring; not for private-data collection or engagement automation.
Extended `5dive` CLI recipes beyond the everyday core — see the `5dive-cli` skill first for spawning/messaging sibling agents and the basic task queue. Use THIS skill for hiring a ready-made persona off the agent market (`5dive market`, `hire --from-market`) or firing one (`5dive fire`), auth recovery (`error.class=auth_required`, `--defer-auth`, device-code login via `agent auth start/poll/submit`), BYO-provider agents (`--provider`), multi-account auth (`5dive account`), declarative fleets and company templates (`5dive up/down/ps/export`, `team import`), hosting a CrewAI crew (`5dive crew`), controlling agents on OTHER registered boxes (`5dive fleet`), recurring/scheduled work (`task add --recurring`, `5dive heartbeat`) and projects (`5dive project add`), building or editing multi-agent loops — a relay with optional human gates (`task loop start`) or a maker→verifier review loop (`task add --verifier`, `task reject`, `5dive loop` LOOP-7 verbs) — decomposing an outcome into a guardrailed task DAG (`5dive goal add`) or a self-steering objective bound to a live metric (`5dive objective`), compiling durable knowledge into the shared wiki (`5dive memory add`), org-chart writes (`5dive org set`), convening a governance vote (`5dive council`), reading fleet health / token burn / the daily standup (`5dive supervisor`, `5dive usage`, `5dive digest`), a machine-readable health check (`5dive doctor --json`, `5dive selfcheck --json`), a task's causal history (`5dive trace`), the current model id per alias (`5dive models`), Telegram/Discord pairing and shared team-bot setup, a delegated GitHub push-for-review (`5dive push`), or the onboarding wizard (`5dive company`).
When the user wants to write, rewrite, or improve marketing copy for any page — including homepage, landing pages, pricing pages, feature pages, about pages, or product pages. Also use when the user says "write copy for," "improve this copy," "rewrite this page," "marketing copy," "headline help," "CTA copy," "value proposition," "tagline," "subheadline," "hero section copy," "above the fold," "this copy is weak," "make this more compelling," or "help me describe my product." Use this whenever someone is working on website text that needs to persuade or convert. For email copy, see emails. For popup copy, see popups. For editing existing copy, see copy-editing.
Parse a local or remote FILE (PDF) into markdown or structured JSON with fastCRW. Use when the source is a file on disk — "parse this PDF", "extract text from this document", "read this report", "convert PDF to markdown". Routing rule: URL → use crw-scrape; file on disk → use crw-parse. Step 5 of the crw workflow ladder.
HTX spot market data — ticker / klines / order book / latest trades / currency and symbol metadata.
Copy-trade any Hyperliquid trader automatically. Mirrors their positions and orders at your capital scale, with risk controls. Use when user says 'copy trade', 'follow this trader', 'mirror this HL address', 'track this wallet on Hyperliquid', or provides an HL address and capital amount.
Use this skill whenever a user specifically requests legal research or Westlaw Deep Research, asks for CoCounsel Legal or cocounsel legal support, or asks a question that requires explaining, analyzing, or synthesizing U.S. law.
Self-evolving skills. Record user feedback, update methodologies and rules. Trigger words: update rules, record feedback, improve skill