flow-nexus-swarm

---

LIBRARY-FIRST PROTOCOL (MANDATORY)

Before writing ANY code, you MUST check:

Step 1: Library Catalog

• Location: .claude/library/catalog.json
• If match >70%: REUSE or ADAPT

Step 2: Patterns Guide

• Location: .claude/docs/inventories/LIBRARY-PATTERNS-GUIDE.md
• If pattern exists: FOLLOW documented approach

Step 3: Existing Projects

• Location: D:\Projects\*
• If found: EXTRACT and adapt

Decision Matrix

Match	Action
Library >90%	REUSE directly
Library 70-90%	ADAPT minimally
Pattern exists	FOLLOW pattern
In project	EXTRACT
No match	BUILD (add to library after)

---

STANDARD OPERATING PROCEDURE

Purpose

Deploy and manage Flow Nexus swarms with topology-aware scaling, secure channel setup, and resilience against cloud-specific failure modes.

Trigger Conditions

• Positive: Flow Nexus deployments, cloud swarm tuning, scaling/rollback drills, secure channel provisioning, cross-region coordination.
• Negative: local-only swarms, non-Flow Nexus requests, or pure prompt shaping (route to prompt-architect).

Guardrails

• Skill-Forge structure-first: ensure SKILL.md, examples/, tests/ exist; add resources/ and references/ or log remediation.
• Prompt-Architect clarity: extract HARD/SOFT/INFERRED constraints (regions, quotas, compliance), use pure English, and declare confidence ceilings.
• Platform safety: enforce identity, network, and secret management policies; register agents; keep hook latency within budgets; define rollback for cluster actions.
• Adversarial validation: simulate region loss, quota exhaustion, and message drops; record evidence and metrics.
• MCP tagging: store swarm runs with WHO=flow-nexus-swarm-{session} and WHY=skill-execution.

Execution Playbook

1. Intent & constraints: capture targets (capacity, latency, compliance) and confirm inferred platform limits.
2. Topology & setup: design swarm layout, provision channels, register agents, and configure autoscaling rules.
3. Deployment plan: stage rollout, health checks, and backoff/retry policies; set observability hooks.
4. Safety nets: pre-mortem failure points, define rollback and isolation controls; validate secrets and access scopes.
5. Validation loop: run adversarial drills (failover, rate limits), measure telemetry, and capture evidence.
6. Delivery: present deployment map, validation results, residual risks, and confidence ceiling.

Output Format

• Deployment objective, constraints, and topology.
• Channel/security setup, autoscaling rules, and rollback paths.
• Validation artifacts (failover, latency, quota tests) and risk log.
• MCP references for persisted notes.
• Confidence: X.XX (ceiling: TYPE Y.YY) - rationale.

Validation Checklist

• Structure-first assets present or ticketed; examples/tests updated or planned.
• Identity/secrets validated; registry and hooks healthy; rollback tested.
• Adversarial and COV runs logged with MCP tags; confidence ceiling stated; English-only output.

Completion Definition

Swarm deployment is complete when topology is live, security and scaling controls are verified, evidence is stored, and remaining risks are owned with next actions.

Confidence: 0.70 (ceiling: inference 0.70) - Flow Nexus orchestration reframed with skill-forge structure and prompt-architect evidence rules for cloud resilience.