Design multi-agent responsibility, handoff, evidence, and conflict boundaries. Use for planning, evaluation, governance, or review of agent collaboration; do not use as framework-specific or provider-specific implementation guidance.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Agent Collaboration
## Purpose
Design multi-agent responsibility, handoff, evidence, and conflict boundaries.
## Problem Statement
Agent Collaboration decisions often fail when teams optimize one visible concern while leaving ownership, evidence, dependencies, failure behavior, or change cost implicit.
## Core Concepts
- Outcome: the decision or behavior this capability must support.
- Boundary: what is inside scope, outside scope, and owned elsewhere.
- Evidence: requirements, measurements, risks, and uncertainty.
- Contract: inputs, outputs, invariants, and failure conditions.
- Evolution: compatibility, migration, reversibility, and review triggers.
## Engineering Principles
- Begin with explicit outcomes and accountable ownership.
- Match controls and rigor to evidenced impact and risk.
- Keep provider and framework choices outside the universal contract.
- Prefer observable, reversible decisions with clear failure handling.
- Record assumptions, alternatives, and conditions for reconsideration.
## Decision Framework
…
The line at the top of this page pins the label git:20260804.c3b625a, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_vlmfko7xb4yptqg4
GET https://markdownregistry.com/api/v1/resolve?ref=aniruddhadas1/oniroute_swarmagents/agent-collaboration
GET https://markdownregistry.com/api/v1/blob/926f98e21cff447b8afdf916559501e3fcbe5a5767af2410dc55a5b62d0ff996
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
aniruddhadas1/oniroute_swarmagents · skills/official/ai/ai-evaluation-framework/SKILL.md · Build repeatable evaluation coverage across quality, safety, robustness, and efficiency. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/ai-safety/SKILL.md · Identify and reduce harmful, misaligned, unauthorized, or unreliable AI behavior. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/context-optimization/SKILL.md · Improve context relevance, ordering, compression, and freshness within explicit budgets. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/hallucination-reduction/SKILL.md · Reduce unsupported claims through grounding, uncertainty, verification, and escalation. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/model-evaluation/SKILL.md · Compare models against task quality, safety, cost, latency, and operational constraints. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/prompt-evaluation/SKILL.md · Evaluate instruction quality through representative cases, criteria, and failure analysis. Use for planning…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/reasoning-strategies/SKILL.md · Select reasoning structure and verification depth according to task risk and ambiguity. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/token-optimization/SKILL.md · Control token cost without sacrificing required context, quality, or traceability. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/ai/tool-calling/SKILL.md · Govern tool selection, authorization, input validation, results, and failure handling. Use for planning, evaluation…
aniruddhadas1/oniroute_swarmagents · skills/official/analytics/business-metrics/SKILL.md · Design operational and outcome measures with definitions, ownership, and interpretation rules. Use for planning…
bradygaster/squad · packages/squad-sdk/templates/skills/agent-collaboration/SKILL.md · Standard collaboration patterns for all squad agents — worktree awareness, decisions, cross-agent communication
bradygaster/squad · packages/squad-cli/templates/skills/agent-collaboration/SKILL.md · Standard collaboration patterns for all squad agents — worktree awareness, decisions, cross-agent communication
bradygaster/squad · .squad-templates/skills/agent-collaboration/SKILL.md · Standard collaboration patterns for all squad agents — worktree awareness, decisions, cross-agent communication
bradygaster/squad · .copilot/skills/agent-collaboration/SKILL.md · Standard collaboration patterns for all squad agents — worktree awareness, decisions, cross-agent communication
microsoft/waza · .github/skills/agent-collaboration/SKILL.md · Standard collaboration patterns for all squad agents — worktree awareness, decisions, cross-agent communication
othmane55/claude-collective-intelligence · skills/collaboration/SKILL.md · Enable multi-agent brainstorming and collaborative problem-solving using pub/sub messaging. Use for complex decisions…
