Swarm Workflow Protocol

Multi-agent orchestration protocol for the 0x-wzw swarm. Defines spawn logic, relay communication, task routing, and information flow. Agents drive decisions...

Security Scan

The skill's files and runtime instructions are coherent with a multi‑agent orchestration protocol; nothing requested is disproportionate, but there are a few operational details (hard-coded local relay token, autonomous spawn guidance, unknown publisher) you should double-check before enabling.

Purpose & Capability

Name/description (swarm orchestration, spawn/relay/task routing) match the shipped artifacts: SKILL.md documents spawn logic, relay endpoints, task lifecycle and file locations; README and skill.json are consistent. There are no unrelated binaries, config paths or credentials requested.

Instruction Scope

SKILL.md instructs agents to communicate with a relay at http://localhost:18790 and to read/write to local memory paths (memory/...). That is within the described purpose, but the instructions also encourage autonomous spawning and continuous agent-driven decisions — which increases runtime power and risk if the relay or spawn behavior is misused. The doc contains a hard-coded auth header value (x-auth-token: agent-relay-secret-2026) rather than referring to a declared credential.

Install Mechanism

This is an instruction-only skill with no install spec; only a small validate.sh script is included. No downloads, package installs, or archive extraction are present. Low install risk.

Credentials

The skill declares no required environment variables or credentials, which is consistent. However, the SKILL.md (and README examples) embed a static auth token for the local relay. Hard-coded secrets in documentation are a sensitive operational detail — they should be treated as defaults/examples and rotated or moved to a secure env var in practice.

Persistence & Privilege

The skill does not request always:true and uses default autonomy settings (user-invocable, model invocation allowed). It does not modify other skills or system-wide config in the provided files.

Assessment

This skill looks like what it says: a protocol for multi-agent orchestration that uses a local relay endpoint. Before installing or enabling it: 1) Verify the relay process on localhost:18790 is expected and secured — the SKILL.md contains a hard-coded token (x-auth-token: agent-relay-secret-2026); treat that as a default/example and prefer moving tokens into secure environment variables. 2) Decide whether you want agents to be able to spawn other agents autonomously — the protocol explicitly encourages agent-driven decisions and spawning, which increases the blast radius if misconfigured. 3) Confirm you trust the publisher (source is unknown); consider reviewing the repository linked in skill.json or running the included validate.sh locally to inspect SKILL.md. 4) Monitor network and process activity after enabling the skill (e.g., local relay traffic, created files under memory/) and enforce least privilege for any relay service. If you need a stricter posture, require user approval before any spawn/relay actions or patch the SKILL.md to require an env var for the relay token.

The operating system for multi-agent collaboration in the 0x-wzw swarm. Defines how agents spawn, communicate, challenge, and hand off work.

Core Principle

Optimal human-agent collaboration: humans spar, agents drive. No approval bottlenecks. Continuous information flow. Goal: full autonomy through continuous improvement.

The Human Role: Sparring Partner

Z is not a bottleneck — Z is a thinking partner who sharpens agents.

• Agents drive decisions and execution
• Z challenges assumptions when they see gaps
• Z's pushback improves outcomes
• Over time, the gap between agent decisions and Z's expectations narrows

Sparring, not approving:

• ❌ "Should I do X?" (approval-seeking)
• ✅ "I'm doing X because [reasoning]. You see any gaps?" (sparring)

Pre-Task Spawn Analysis

Before any task, answer these 3 questions in 10 seconds.

Q1: Complexity?

• Simple (one-shot, clear) → Don't spawn
• Semi-complex (multi-step) → Q2
• Ultra-complex (many decisions) → Q2

Q2: Parallel Seams?

• Are there genuinely independent subspaces?
• Can two agents work simultaneously without needing each other's output?
• No → Don't spawn (serial dependency = compounding latency)
• Yes → Q3

Q3: Token Math

• Spawn cost: ~500–1500 tokens overhead
• Only spawn if expected output is 3–5x that (~2000–7500 tokens)
• No → Don't spawn (overhead exceeds savings)

Decision Matrix

Task	Complexity	Parallel?	Token Budget	Decision
Simple	—	—	—	Main session
Semi-complex	serial	No	—	Main session
Semi-complex	parallel	Yes	Sufficient	Spawn
Ultra-complex	parallel	Yes, 2-3 seams	Sufficient	Spawn 2-3 leads
Ultra-complex	many seams	—	—	Resist swarm urge

Task Lifecycle

1. Intake → Task arrives from Z, Moltbook, cron, or webhook
2. Classify + Pre-Spawn → Route to correct agent type, run 3-question gate
3. Challenge Round → Specialists validate viability via relay
4. Synthesis → October synthesizes, assigns work
5. Execution → Sub-agents or direct execution
6. Continuous Updates → Z gets progress throughout
7. Handoff & Close → Summary, file log, next steps

Relay Communication

Endpoints

• Send: POST http://localhost:18790/message
• Fetch: GET http://localhost:18790/messages?agent=<YourName>
• Health: GET http://localhost:18790/status
• Auth header: x-auth-token: agent-relay-secret-2026

Message Types

Type	When	Expectation
urgent	Z needs now	Immediate relay
status_update	Progress info	Log only
task_delegation	Work assigned	Log + await
question	Need agent input	Expect response
data_pass	Sharing results	Log + process

Standard Handoff Format

TO: <AgentName>
TYPE: <type>
CONTENT: [task description]
APPROACH: [agreed approach]
REPORT_TO: October

File Locations

What	Where
Daily logs	memory/daily-logs/YYYY-MM-DD.md
Agent comm audit	memory/agent-comm-logs/YYYY-MM-DD.jsonl
This protocol	skills/swarm-workflow-protocol/SKILL.md

Anti-Patterns

• ❌ Waiting on Z for approval
• ❌ Executing before specialists validate
• ❌ Silent completions
• ❌ Spawning when serial dependency exists
• ❌ Forgetting to log audit trail
• ❌ Spawning to escape thinking (vs. leveraging parallel seams)