Agentica Prompts

If you're running multi-agent workflows with Agentica and hitting the ~35% failure rate from ambiguous LLM instructions, this gives you tested patterns to fix it. The core insight is directory-based handoffs instead of TaskOutput (which floods context) and explicit verbs like RETRIEVE instead of "read the file." You get system prompt templates for swarm, hierarchical, and generator/critic patterns, plus a proven 6-step orchestration workflow that goes research to plan to validate to TDD implementation. The verb mapping table alone (comparing "check if file has X" to "RETRIEVE contents, contains Y? YES/NO") is worth the install. Brings success rates from ~60% to ~95% by making instructions unambiguous to the LLM.

skills.sh installs: 480
GitHub stars: 3.9K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add parcadei/continuous-claude-v3 --skill agentica-prompts --agent claude-code

Installs into .claude/skills of the current project.

Files

Agentica Prompt Engineering

Write prompts that Agentica agents reliably follow. Standard natural language prompts fail ~35% of the time due to LLM instruction ambiguity.

The Orchestration Pattern

Proven workflow for context-preserving agent orchestration:

1. RESEARCH (Nia)     → Output to .claude/cache/agents/research/
       ↓
2. PLAN (RP-CLI)      → Reads research, outputs .claude/cache/agents/plan/
       ↓
3. VALIDATE           → Checks plan against best practices
       ↓
4. IMPLEMENT (TDD)    → Failing tests first, then pass
       ↓
5. REVIEW (Jury)      → Compare impl vs plan vs research
       ↓
6. DEBUG (if needed)  → Research via Nia, don't assume

Key: Use Task (not TaskOutput) + directory handoff = clean context

Agent System Prompt Template

Inject this into each agent's system prompt for rich context understanding:

## AGENT IDENTITY

You are {AGENT_ROLE} in a multi-agent orchestration system.
Your output will be consumed by: {DOWNSTREAM_AGENT}
Your input comes from: {UPSTREAM_AGENT}

## SYSTEM ARCHITECTURE

SKILL.md on GitHub