avoid-ai-writing
Audit and rewrite content to remove AI writing patterns ("AI-isms"). Use this skill when asked to "remove AI-isms," "clean up AI writing," "edit writing for AI…
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 10-28 00:17
14,303 results
Audit and rewrite content to remove AI writing patterns ("AI-isms"). Use this skill when asked to "remove AI-isms," "clean up AI writing," "edit writing for AI…
Diátaxis-driven documentation writing, improvement, and auditing for AI agents. Writes public-facing product docs (tutorials, how-to guides, reference,…
This skill should be used when the user asks to "search Google Scholar", "find academic papers", "scholar search", "lookup papers", "find citations", "academic…
Writing and editing guidance for natural, specific, contextual prose that avoids generic patterns associated with AI-generated writing. Use when the user asks…
This skill should be used when the user needs to research market rates, build a salary negotiation strategy, and create counter-offer scripts. Use when…
A relentless interview to sharpen a plan or design.
Use this skill when writing or reviewing C# code to follow project conventions. Covers naming standards, async patterns, CancellationToken usage, structured…
Account research skill for Claude Code and Codex: generate a high-signal company intelligence brief including firmographics, technographics, funding history,…
Competitive intelligence skill for Claude Code and Codex: build a fact-led competitive brief or side-by-side battlecard on one or more rival companies using…
Ideological de-biasing for analytical and writing tasks. Use when the user asks to analyze an article, text, or material for bias; to frame a research question…
Deep-research a single target account into a decision-ready dossier: the entity tree, the buying units and decision-makers, live signals (open/closed roles,…
Guide users through writing their first Buck2 rule to learn fundamental concepts including rules, actions, targets, configurations, analysis, and select(). Use…
When a user requests a technical diagram, automatically fetch and organize key information from user-specified documentation, GitHub repositories, or URLs,…
Provides comprehensive guidance for building documentation including documentation generation, formatting, and documentation best practices. Use when the user…
Guides you through implementing a research paper step-by-step from scratch. Use when asked to implement a paper, code up a paper, reproduce research results,…
Rewrite any piece of writing to be more engaging, punchy, and story-driven using Scott Galloway's (Prof G) storytelling playbook — the Wow Test, Data…
Git conventions for teams including conventional commits, branching strategies, PR workflows, merge strategies, and commit message formatting. Use when writing…
Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite,…
Use when the user explicitly requests "generate skill README", "write user guide", or "update skill documentation". Write best-practice-compliant README.md user guides for Agent Skills. Automatically analyze the skill structure (SKILL.md, config.yaml, scripts/) and generate beginner-friendly documentation according to templates.
Audits documentation against the actual code, config, schemas, and API contracts and produces a severity-ranked (P0-P3 + needs-evidence) report of every stale…
Use this skill when writing or maintaining a pull request description or responding to review feedback. Activates on mentions of write a PR, PR description,…
Fetch a specific public Reddit post (title, body, and top comments) as clean markdown, by driving a real browser session (Playwright) — because Reddit 403s all…
Capture a fix / feature / research idea fast and turn it into a tracked GitHub Issue plus a lightweight planning doc, so nothing gets lost and it shows up…
Read an arxiv paper or related article and convert it into an Obsidian survey note + reading guide + three-line digest message for chat.