llm-communication
Write effective LLM prompts, commands, and agent instructions. Goal-oriented over step-prescriptive. Role + Objective + Latitude pattern. Use when writing prompts, designing agents, building Claude Code commands, or reviewing LLM instructions. Keywords: prompt engineering, agent design, command writing.
npx skills add majiayu000/claude-skill-registry --skill llm-communication --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Talking to LLMs This skill helps you write effective prompts, commands, and agent instructions. ## Core Principle LLMs are intelligent agents, not script executors. Talk to them like senior engineers. ## Anti-Patterns ### Over-Prescriptive Instructions Bad: ``` Step 1: Run `sentry-cli issues list --status unresolved` Step 2: Parse the JSON output Step 3: For each issue, calculate priority score using formula... Step 4: Select highest priority issue Step 5: Run `git log --since="24 hours ago"` ...700 more lines ``` This treats the LLM like a bash script executor. It's brittle, verbose, and removes the LLM's ability to adapt. ### Excessive Hand-Holding Bad: ``` If the user says X, do Y. If the user says Z, do W. Handle edge case A by doing B. Handle edge case C by doing D. ``` You can't enumerate every case. Trust the LLM to generalize. ### Defensive Over-Specification Bad: ``` IMPORTANT: Do NOT do X. WARNING: Never do Y. CRITICAL: Always remember to Z. ``` If you need 10 warnings, your instruction is probably wrong. ## Good Patterns ### State the Goal, Not the Steps Good: ``` Investigate production errors. Check all available observability (Sentry, Vercel, logs). Correlate with gi
- Core Principle
- Anti-Patterns
- Over-Prescriptive Instructions
- Excessive Hand-Holding
- Defensive Over-Specification
- Good Patterns
- State the Goal, Not the Steps
- Provide Context, Not Constraints
- Trust Recovery
- Role + Objective + Latitude
- When Writing Claude Code Commands
- When Building Agentic Systems
- The Test
- Remember
What does the llm-communication skill do?
Write effective LLM prompts, commands, and agent instructions. Goal-oriented over step-prescriptive. Role + Objective + Latitude pattern. Use when writing prompts, designing agents, building Claude Code commands, or reviewing LLM instructions. Keywords: prompt engineering, agent design, command writing.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill llm-communication --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.