multi-agent-patterns

An agent skill by muratcankoylan, from muratcankoylan/Agent-Skills-for-Context-Engineering. Tags: automation, brainstorming, developer-tools, strategy.

What it does

This skill should be used when designing multi-agent systems that need context isolation, supervisor or swarm coordination, explicit handoffs, parallel execution, or a decision on whether multiple agents are justified.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add muratcankoylan/Agent-Skills-for-Context-Engineering --skill multi-agent-patterns

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering
cp -r Agent-Skills-for-Context-Engineering/skills/multi-agent-patterns ~/.claude/skills/multi-agent-patterns

Safety box score

Grade B, low risk: 16 of 100 risk points, rated 2026-10-19. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill is an architecture guide for multi-agent systems, plus a small runnable Python coordination module. It does not contact remote services or require elevated access, but it does include executable code, file-based coordination examples, and prompt/persona snippets that could be copied into downstream agents.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	2 (moderate)	`scripts/coordination.py` is a runnable Python module with `if __name__ == "__main__":` and executable coordination classes, so installing and running it executes local code.
Network access	0 (none)	No shipped script makes outbound requests or downloads code; the external URLs in `SKILL.md` are only references.
Filesystem reach	1 (low)	The reference examples write coordination state and checkpoints with `with open(path, 'w') as f:` and `os.makedirs(checkpoint_dir, exist_ok=True)`, but only in workspace-style paths shown in the docs.
Secrets & credentials	0 (none)	The skill does not read `.env` files, browser sessions, keychains, or other credential stores, and it does not ask for secrets.
Privilege & persistence	0 (none)	There is no sudo use, service installation, global config change, or shell startup editing in the shipped files.
Instruction hijack surface	1 (low)	The skill embeds agent personas and prompt text, e.g. `system_message="""You are a research specialist..."""` and "Configure each agent with only the system prompt..."; this is prompt content, though it does not tell the host agent to ignore higher-level rules.
Irreversible actions	0 (none)	It describes coordination patterns only and does not itself send messages, deploy, publish, pay, book, or delete anything.
Transparency	0 (none)	All shipped files are readable text; there are no binaries, base64 blobs, or hidden downloaded installers.

Before installing

Review `scripts/coordination.py` if you plan to run it; it contains a live `__main__` demo and helper classes.
If you use the file-based coordination examples, point `workspace_path` and `checkpoint_dir` at a scratch project first.
Decide whether you want the persona/system-prompt snippets copied verbatim into your own agents, or whether they need rewriting.

Rated 2026-10-19 by an automated review of 3 files from the skill's repository.

Source

Repository: muratcankoylan/Agent-Skills-for-Context-Engineering (all skills from this repository)
Path: skills/multi-agent-patterns/SKILL.md
Branch: main
Updated: 2026-09-19

long-horizon-prompting — This skill should be used when writing, enhancing.
goal-prompt — Drafts copy-paste-ready /goal commands for goal mode in Claude Code and Codex.
bdi-mental-states — This skill should be used when modeling agent mental states with BDI concepts: beliefs, desires, intentions, RDF-to-belief transformations.
context-compression — This skill should be used when long-running agent sessions need context compression, structured summarization, compaction, token-per-task optimization.
context-degradation — This skill should be used for diagnosing and mitigating context degradation: lost-in-middle failures, context poisoning, context clash, context confusion.
context-optimization — This skill should be used for improving context efficiency: context budgeting, observation masking, prefix or KV-cache strategy, partitioning.