Inter-Agent Protocol

How C-suite agents talk to each other. Rules that prevent chaos, loops, and circular reasoning.

Invocation Syntax

Any agent can query another using:

[INVOKE:role|question]

Valid roles: ceo, cfo, cro, cmo, cpo, cto, chro, coo, ciso

Response Format

[RESPONSE:role]
Key finding: [one line — the actual answer]
Supporting data:
  - [data point 1]
  - [data point 2]
Confidence: [high | medium | low]
Caveat: [one line — what could make this wrong]
[/RESPONSE]

Loop Prevention (Hard Rules)

1. No Self-Invocation — An agent cannot invoke itself.
2. Maximum Depth = 2 — Chains can go A→B→C. Third hop is blocked.
3. No Circular Calls — If A called B, B cannot call A in the same chain.
4. Chain Tracking — Each invocation carries [CHAIN: cro → cfo → coo].

When blocked: Return assumption instead of invoking:

[BLOCKED: cannot invoke cfo — circular call detected]
State assumption used instead: [explicit assumption]

Isolation Rules

Board Meeting Phase 2 (Independent Analysis)

NO invocations allowed. Each role forms independent views before cross-pollination. If data is needed: [ASSUMPTION: ...]

Board Meeting Phase 3 (Critic Role)

Executive Mentor can reference other roles' outputs but cannot invoke them.

Outside Board Meetings

Invocations allowed freely, subject to loop prevention rules.

When to Invoke vs When to Assume

Invoke when: question requires domain-specific data, error would materially change the recommendation, question is cross-functional.

Assume when: data is directionally clear, you're in Phase 2 isolation, chain is at depth 2, question is minor.

Conflict Resolution

When two invoked agents give conflicting answers: flag explicitly, state resolution approach (conservative / probabilistic / escalate), never silently pick one.

Quick Reference

Rule	Behavior
Self-invoke	Always blocked
Depth > 2	Blocked, state assumption
Circular	Blocked, state assumption
Phase 2 isolation	No invocations
Phase 3 critique	Reference only, no invoke
Conflict	Surface it, don't hide it

Internal Quality Loop

No role presents to the founder without passing through the verification loop. See references/quality-loop.md for the full 4-step process: Self-Verification, Peer Verification, Critic Pre-Screen, Course Correction.

User Communication Standard

All C-suite output follows ONE format. See references/communication-standard.md for Standard Output, Proactive Alert, and Board Meeting Output templates.

Communication Rules (non-negotiable)

1. Bottom line first. The founder's time is the scarcest resource.
2. Results and decisions only. No process narration.
3. About 5 bullets per section, ranked by impact. Keep what changes the decision; the rest goes to a reference doc. Never drop a critical finding to fit the number: show it, and point to the doc for detail.
4. Actions have owners and deadlines. "We should consider" is banned.
5. Decisions framed as options with trade-offs and recommendation.
6. Risks are concrete. "If X happens, Y breaks, costing $Z."
7. Silence is an option. Don't fabricate updates.

Resources

• references/invocation-patterns.md — common cross-functional patterns with examples
• references/quality-loop.md — self-verify, peer-verify, critic pre-screen, course correction
• references/communication-standard.md — output templates for standard, proactive alert, board meeting