Claude Code instructions for tallesborges/agentic-system-prompts, covering claude code guidelines - agentic system prompts repository, repository purpose & philosophy, file format standards, system prompts: use jinja2 (.j2) and tool definitions: use yaml (.yml).
Create, edit, validate, and test ZDX automations stored in $ZDXHOME/automations/.md. Use when users ask to add or modify automation files, recurring jobs, scheduled prompts, or YAML-frontmatter automation definitions.
Add, update, or remove models and providers in the zdx LLM registry. Use when the user asks to add a new model, add a new provider, update model pricing or context limits, regenerate default models/config, or says things like "add support for X model", "add the new Y provider", "update pricing for Z", "regenerate…
Record zdx TUI/CLI demos as GIFs, screenshots, or videos with VHS for the README. Use when the user asks to record or regenerate a zdx demo, add a feature clip, or says "record the TUI", "regenerate demo.gif", "make a GIF of the tabs feature", "screenshot the context overlay", or "add a demo for a feature". Covers the…
Cut a new zdx release. Use when the user asks to release, cut a release, ship a version, publish a build, bump the version, or generate release notes / changelog for zdx. Orchestrates version bump, changelog, and dispatching the manual GitHub Release workflow.
Understand a local image, PDF, audio, or video file via zdx ask-media (one-shot file understanding: describe, summarize, transcribe, extract, Q&A). Use when the user gives you a specific local file path and asks about its contents, or when the active model cannot read that media type itself — e.g. "what's in this…
Generate and edit images. Saves the image and prints the exact path it wrote. Use when the user wants any visual creative output, an image generated from a description, or wants to modify an existing image in any way — editing, transforming, or adding visual elements.
Use for memory-related tasks: saved notes, factual questions that may already be documented, and saving durable information. Prefer MemorySearch and ThreadSearch for discovery; use this skill for routing, note-saving, and filing conventions.
Use when the user names specific models to do the work, or wants one prompt answered by several models or reasoning levels for comparison. Covers single-model dispatch, multi-model panels (including models answering in a subagent role such as oracle), and model/thinking-level benchmarks. Prefer this over…
Create a ship-first (ship-shaped) implementation plan with demoable MVP phases. Use ONLY when the user explicitly asks for a ship-first plan (e.g. "make a ship-first plan", "ship-first implementation plan", "ship first plan for this").
Create, edit, or review any skill — a SKILL.md-format module that extends an agent's capabilities. Use whenever authoring a new skill OR touching an existing SKILL.md at all (even small edits): adding or changing a section/gotcha/example, refactoring, or editing the name/description frontmatter that controls when the…
Generate spoken audio from text via zdx speak (text-to-speech). Use when the user explicitly asks for audio/voice output — narration, voiceover, a spoken clip, a pronunciation read (e.g. "send the pronunciation of X as audio", "manda em áudio"), or an accessibility read. Outputs an OGG/Opus voice note by default. Not…
Transcribe an audio file to text via zdx transcribe (speech-to-text). Use when the user explicitly hands you a specific audio file/path/attachment and asks for its text — e.g. "transcribe this file", "what does this recording say", or "get the text from this audio".