C-Level Advisory ProtocolAgentsCommunication

Agent Protocol

Inter-agent communication protocol for C-suite agent teams.

$skillfish add ~skills/agent-protocol

Capabilities Real Examples Get Started Models & Context

What This Skill Can Do

Concrete capabilities you get when you install this skill.

Design multi-agent debate protocols with structured roles and decision rules

Facilitate structured disagreement between AI agents on strategic questions

Build consensus workflows across parallel agent workstreams

Define agent handoff protocols and context transfer between sessions

Design escalation paths and human checkpoints in multi-agent workflows

Create structured briefing formats for sequential multi-agent processes

Real Conversations

See how the Agent Protocol skill responds to actual requests. These are representative examples of what your AI agent will produce.

You

I want to run a weekly strategic review where CEO advisor, CFO advisor, and CTO advisor each contribute their perspective on the same company update. How do I structure the agent handoffs so the insights build on each other instead of being siloed?

Agent

Multi-agent strategic reviews require a defined handoff protocol so each advisor builds on the previous one rather than starting from scratch.

The Sequential Briefing Protocol:

Step 1 — Shared Context Packet (you provide once)

[COMPANY CONTEXT]
[THIS WEEK'S UPDATE: metrics, events, decisions pending]
[QUESTION FOR EACH ADVISOR]

Step 2 — Ordered Advisor Sequence

Run in this order (finance informs strategy, strategy informs technology):

CFO Advisor → financial read on the week's update
- Outputs: Cash position implication, forecast adjustment, financial risk flags
- Pass to next: "CFO noted [X financial constraint / opportunity]. Factor this in."
CEO Advisor → strategic synthesis incorporating CFO perspective
- Input: Week's update + CFO output
- Outputs: Strategic priorities, stakeholder implications, decisions needed
- Pass to next: "CEO strategic frame is [X]. Here's what this means for technical priorities."
CTO Advisor → technical and operational response
- Input: Week's update + CEO strategic frame
- Outputs: Technical feasibility, engineering capacity constraints, build vs. buy recommendations

Handoff Message Template:

Previous advisor output:
[Paste CFO/CEO output]

Your role: [CTO/CEO/CFO]
Your specific question this week: []

Build on the previous perspective, note where you agree or disagree, and add your domain's lens.

Anti-pattern to avoid: Running all three advisors on the same prompt independently — you get parallel, unintegrated advice instead of a genuine board conversation.

What's your typical weekly review cadence — do you want a template for the weekly packet?

You

CEO advisor says hire a VP of Sales now. CFO advisor says we can't afford it given 11 months of runway. CTO advisor is neutral but wants engineering headcount instead. How does the agent protocol handle genuine disagreement?

Agent

Conflicting advisor recommendations are a feature, not a bug — they reflect a real organizational tension that needs structured resolution, not a tiebreaker.

The Conflict Resolution Protocol:

Step 1 — Articulate the underlying disagreement The surface conflict (hire vs. don't hire) masks a deeper disagreement about:

Risk tolerance: Is 11 months enough runway to absorb a high-burn hire?
Causal theory: Does revenue growth precede or follow a VP of Sales hire?
Priority stack: Sales capacity vs. engineering capacity — which is the current bottleneck?

Step 2 — Request a structured debate

Prompt format:

"CFO and CEO advisors have conflicting recommendations on VP of Sales hire.
Run a structured debate:
- CEO: strongest case FOR the hire
- CFO: strongest case AGAINST or for an alternative
- Synthesis: What additional information would resolve this disagreement?"

Step 3 — Identify the deciding variable Good advisor disagreements usually hinge on one key assumption. In this case:

If pipeline coverage is >3x and win rate is stable → revenue case for VP is stronger
If pipeline is thin → VP won't help; the problem is lead generation, not conversion
If runway extends to 18+ months with a good Q4 → CFO objection weakens

Step 4 — Decision with recorded rationale Use the Decision Logger to capture: what was the disagreement, what resolved it, what was decided and why.

For your specific situation: The resolution question is: what's the minimum viable sales investment that doesn't put runway below 9 months? Often the answer is a senior AE at 60% of VP cost as a 90-day test.

What's your current pipeline coverage ratio?

Get Started

How to install and use this skill in your preferred environment.

Skills are designed for AI coding agents (Claude Code, Cursor, Windsurf) and IDE-based workflows where the agent can read files, run scripts, and act on your codebase. Web-based AI can use the knowledge and frameworks, but won't have tool access.

Models & Context

Which AI models and context windows work best with this skill.

Recommended Models

Best

Claude Opus 4 Claude Sonnet 4 GPT-4.1 Gemini 2.5 Pro Grok 3 Kimi K2

Good

Claude Haiku 4.5 GPT-4.1 mini Gemini 2.5 Flash Grok 3 mini

Larger models produce more detailed, production-ready outputs.

Context Window

This skill's SKILL.md is typically 3–10 KB — fits in any modern context window.

8K Skill only

32K+ Skill + conversation

100K+ Skill + references + codebase

All current frontier models (Claude, GPT, Gemini) support 100K+ context. Use the full window for complex multi-service work.

Pro tips for best results

1

Be specific

Include numbers — users, budget, RPS — so the skill can size the architecture.

2

Share constraints

Compliance needs, team size, and existing stack all improve the output.

3

Iterate

Start with a high-level design, then ask follow-ups for IaC, cost analysis, or security review.

4

Combine skills

Pair with companion skills below for end-to-end coverage.

Ready to try Agent Protocol?

Install the skill and start getting expert-level guidance in your workflow — any agent, any IDE.