calionauta

[Cali] Generate and maintain project AGENTS.md files using semantic analysis. Triggers when: a project has no AGENTS.md, user asks to 'create AGENTS.md', 'generate agents md', 'setup agents md', or when starting work on a project that lacks one. Also triggers when user asks to 'check if AGENTS.md is stale', 'update agents md', or when evolving an existing AGENTS.md. Covers: initial generation from codebase analysis, sem-based staleness detection, pre-commit hooks, and CI guards. Always offer hook installation when invoked for any reason.

[Cali] Validate project AGENTS.md files against best practices and the canonical template. Use when: user says 'validate agents md', 'check agents md quality', 'audit my agents md', or after creating/updating AGENTS.md. Checks: structure, size, content quality, template compliance, and provides fix recommendations. Can offer automatic corrections via ask tool.

[Cali] Tool priority rules for code navigation: Cymbal-first, FFF fallback, batch guidance for token efficiency, and semantic diff workflow. Activates on any file-task within a Git repository to guide agent tool selection.

[Cali] Use this skill whenever the user wants to analyze a codebase and generate a detailed product spec document. Triggers for: "analyze this codebase", "generate a spec", "document this project", "reverse engineer this app", "what does this system do", "create product documentation", "map out the features", "extract product rules from code", "what does this codebase do", "generate product spec from code", "document this app", "spec out this repo", "read this code and tell me what it does", "create a product spec", "reverse engineer the product", "what features does this app have", "document the flows", "extract the prompts from this code", or any request to extract product knowledge from source code. Supports any input method: file uploads, pasted code, filesystem paths, or zip archives. Always use this skill when the user shares code AND wants to understand what the product does — even if they don't explicitly say "spec" or "documentation".

[Cali] Go web stack: Datastar, Templ, DaisyUI, NATS, PocketBase/SQLite, GoAI LLM, Zenflow agents. Scaffold, real-time, hypermedia, auth, DB, embeddings, voice AI, AI agents, durable workflows, queues. Queues use NATS JetStream or goqite. Workflows use go-workflows, turbine, ebind, dagnats, or Hatchet per decision tree below.

Use this skill for any Go (Golang) backend task — writing services, APIs, CLI tools, concurrent code, or reviewing/refactoring existing Go code. Triggers on: any .go file, mention of goroutines, channels, mutexes, Go modules, or requests to build backend systems in Go. Also triggers when setting up linters, running tests, managing Go dependencies, or running the app locally during development.

Builds reactive web applications with StarHTML, a Python-first framework over Datastar. Use when writing StarHTML components, signals, event handlers, reactive attributes, conditional helpers, CSS classes, computed signals, HTTP actions, SSE endpoints, or plugins (persist, scroll, resize, drag, canvas, motion, markdown, katex, mermaid, split, nodegraph). For UI components, use StarUI (shadcn/ui for Python). After generating any StarHTML file, run `starhtml-check <file.py>` to validate.

Adds real-time broadcasting, multi-user collaboration, and database persistence to StarHTML applications. Use when implementing chat systems, collaborative editors, presence tracking, typing indicators, or any feature requiring synchronized state across clients. Supports core StarStream (broadcasting), PocketBase (persistence), and Loro (CRDT) plugins.

Advertising strategies for online placement based on the Transtheoretical Model of Change. Covers the 5 stages of audience awareness and ad categories for each stage.

Business model creativity triggers for reducing costs and generating revenue. Based on Jobs to Be Done frameworks from Strategyn to adapt and experiment with new models.

[stelow] Structural critique for codebases. Accepts a directory of source code and evaluates architecture, data flow, API contracts, performance, theming, responsive patterns, and AI slop in code — then generates a classified gap report. Part of stelow (pre-implementation codebase review) but usable standalone. For codebases with visual UI, use stelow-product-ux-critique instead.

[stelow] Self-contained coding standards for product planning. Universal principles (KISS, DRY, LoB, SoC, Fail Fast, YAGNI, file/function size limits). Use when generating or reviewing code within any product planning context.

[stelow] Complete expert in the short-cycle product learning method, based on the e-book "Guide to Creating Products with Short Learning Cycles" by calionauta. Use this skill whenever the user asks about: idea validation, customer discovery, product experimentation, jobs to be done, pricing strategies, business models, marketing channels, early adopters, MVP, product-market fit, pre-sales, launch strategies, product innovation, product lifecycle, building customer trust, marketplace, open source as strategy, or any concept related to creating and validating digital products with low risk and fast learning. This skill contains ALL the method details — do not skip any detail.

[stelow] Strategic product evolution and innovation exploration inspired by evolutionary systems thinking, novelty search, stepping-stones theory, and adaptive product ecosystems. Activate when: - product strategy feels too objective-driven or roadmap-rigid - teams are prematurely optimizing toward fixed outcomes - exploring innovation, experimentation, or long-term adaptability - evaluating product direction beyond short-term metrics - prioritization discussions feel shallow or disconnected - teams need better framing for experimentation and discovery - discussing ecosystems, platforms, adaptability, or emergent behavior - exploring feature prioritization through connections and optionality - discussing product resilience, flexibility, or future evolution

[stelow] Post-implementation execution critique: verify scope completion, implementation quality, NFR coverage, edge cases, doc/tests, produce a gap registry with decision matrix. Supports 4 modes: workflow (spec-tech.md), plan (spec-product.md), context (dir/URL), and standalone (auto-detects via sem diff + git). Falls back to git diff when sem is not installed.

Product health monitoring through signals in tension. Monitor both effectiveness and user well-being by observing tension between success signals and counterbalance signals, focusing on removing unwanted side effects.