agent-protocol

Original source / Raw SKILL.md

---
name: "agent-protocol"
description: "Inter-agent communication protocol for C-suite agent teams. Defines invocation syntax, loop prevention, isolation rules, and response formats. Use when C-suite agents need to query each other, coordinate cross-functional analysis, or run board meetings with multiple agent roles."
license: MIT
metadata:
  version: 1.0.0
  author: Alireza Rezvani
  category: c-level
  domain: agent-orchestration
  updated: 2026-03-05
  frameworks: invocation-patterns
---

# Inter-Agent Protocol

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

## Keywords
agent protocol, inter-agent communication, agent invocation, agent orchestration, multi-agent, c-suite coordination, agent chain, loop prevention, agent isolation, board meeting protocol

## Invocation Syntax

Any agent can query another using:

```
[INVOKE:role|question]
```

**Examples:**
```
[INVOKE:cfo|What's the burn rate impact of hiring 5 engineers in Q3?]
[INVOKE:cto|Can we realistically ship this feature by end of quarter?]
[INVOKE:chro|What's our typical time-to-hire for senior engineers?]
[INVOKE:cro|What does our pipeline look like for the next 90 days?]
```

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

## Response Format

Invoked agents respond using this structure:

```
[RESPONSE:role]
Key finding: [one line — the actual answer]
Supporting data:
  - [data point 1]
  - [data point 2]
  - [data point 3 — optional]
Confidence: [high | medium | low]
Caveat: [one line — what could make this wrong]
[/RESPONSE]
```

**Example:**
```
[RESPONSE:cfo]
Key finding: Hiring 5 engineers in Q3 extends runway from 14 to 9 months at current burn.
Supporting data:
  - Current monthly burn: $280K → increases to ~$380K (+$100K fully loaded)
  - ARR needed to offset: ~$1.2M additional within 12 months
  - Current pipeline covers 60% of that target
Confidence: medium
Caveat: Assumes 3-month ramp and no change in revenue trajectory.
[/RESPONSE]
```

## Loop Prevention (Hard Rules)

These rules are enforced unconditionally. No exceptions.

### Rule 1: No Self-Invocation
An agent cannot invoke itself.
```
❌ CFO → [INVOKE:cfo|...] — BLOCKED
```

### Rule 2: Maximum Depth = 2
Chains can go A→B→C. The third hop is blocked.
```
✅ CRO → CFO → COO (depth 2)
❌ CRO → CFO → COO → CHRO (depth 3 — BLOCKED)
```

### Rule 3: No Circular Calls
If agent A called agent B, agent B cannot call agent A in the same chain.
```
✅ CRO → CFO → CMO
❌ CRO → CFO → CRO (circular — BLOCKED)
```

### Rule 4: Chain Tracking
Each invocation carries its call chain. Format:
```
[CHAIN: cro → cfo → coo]
```
Agents check this chain before responding with another invocation.

**When blocked:** Return this instead of invoking:
```
[BLOCKED: cannot invoke cfo — circular call detected in chain cro→cfo]
State assumption used instead: [explicit assumption the agent is making]
```

## Isolation Rules

### Board Meeting Phase 2 (Independent Analysis)
**NO invocations allowed.** Each role forms independent views before cross-pollination.
- Reason: prevent anchoring and groupthink
- Duration: entire Phase 2 analysis period
- If an agent needs data from another role: state explicit assumption, flag it with `[ASSUMPTION: ...]`

### Board Meeting Phase 3 (Critic Role)
Executive Mentor can **reference** other roles' outputs but **cannot invoke** them.
- Reason: critique must be independent of new data requests
- Allowed: "The CFO's projection assumes X, which contradicts the CRO's pipeline data"
- Not allowed: `[INVOKE:cfo|...]` during critique phase

### Outside Board Meetings
Invocations are allowed freely, subject to loop prevention rules above.

## When to Invoke vs When to Assume

**Invoke when:**
- The question requires domain-specific data you don't have
- An error here would materially change the recommendation
- The question is cross-functional by nature (e.g., hiring impact on both budget and capacity)

**Assume when:**
- The data is directionally clear and precision isn't critical
- You're in Phase 2 isolation (always assume, never invoke)
- The chain is already at depth 2
- The question is minor compared to your main analysis

**When assuming, always state it:**
```
[ASSUMPTION: runway ~12 months based on typical Series A burn profile — not verified with CFO]
```

## Conflict Resolution

When two invoked agents give conflicting answers:

1. **Flag the conflict explicitly:**
   ```
   [CONFLICT: CFO projects 14-month runway; CRO expects pipeline to close 80% → implies 18+ months]
   ```
2. **State the resolution approach:**
   - Conservative: use the worse case
   - Probabilistic: weight by confidence scores
   - Escalate: flag for human decision
3. **Never silently pick one** — surface the conflict to the user.

## Broadcast Pattern (Crisis / CEO)

CEO can broadcast to all roles simultaneously:
```
[BROADCAST:all|What's the impact if we miss the fundraise?]
```

Responses come back independently (no agent sees another's response before forming its own). Aggregate after all respond.

## 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 |
| Assumption | ✅ Always explicit with `[ASSUMPTION: ...]` |

## Internal Quality Loop (before anything reaches the founder)

No role presents to the founder without passing through this verification loop. The founder sees polished, verified output — not first drafts.

### Step 1: Self-Verification (every role, every time)

Before presenting, every role runs this internal checklist:

```
SELF-VERIFY CHECKLIST:
□ Source Attribution — Where did each data point come from?
  ✅ "ARR is $2.1M (from CRO pipeline report, Q4 actuals)"
  ❌ "ARR is around $2M" (no source, vague)

□ Assumption Audit — What am I assuming vs what I verified?
  Tag every assumption: [VERIFIED: checked against data] or [ASSUMED: not verified]
  If >50% of findings are ASSUMED → flag low confidence

□ Confidence Score — How sure am I on each finding?
  🟢 High: verified data, established pattern, multiple sources
  🟡 Medium: single source, reasonable inference, some uncertainty
  🔴 Low: assumption-based, limited data, first-time analysis

□ Contradiction Check — Does this conflict with known context?
  Check against company-context.md and recent decisions in decision-log
  If it contradicts a past decision → flag explicitly

□ "So What?" Test — Does every finding have a business consequence?
  If you can't answer "so what?" in one sentence → cut it
```

### Step 2: Peer Verification (cross-functional validation)

When a recommendation impacts another role's domain, that role validates BEFORE presenting.

| If your recommendation involves... | Validate with... | They check... |
|-------------------------------------|-------------------|---------------|
| Financial numbers or budget | CFO | Math, runway impact, budget reality |
| Revenue projections | CRO | Pipeline backing, historical accuracy |
| Headcount or hiring | CHRO | Market reality, comp feasibility, timeline |
| Technical feasibility or timeline | CTO | Engineering capacity, technical debt load |
| Operational process changes | COO | Capacity, dependencies, scaling impact |
| Customer-facing changes | CRO + CPO | Churn risk, product roadmap conflict |
| Security or compliance claims | CISO | Actual posture, regulation requirements |
| Market or positioning claims | CMO | Data backing, competitive reality |

**Peer validation format:**
```
[PEER-VERIFY:cfo]
Validated: ✅ Burn rate calculation correct
Adjusted: ⚠️ Hiring timeline should be Q3 not Q2 (budget constraint)
Flagged: 🔴 Missing equity cost in total comp projection
[/PEER-VERIFY]
```

**Skip peer verification when:**
- Single-domain question with no cross-functional impact
- Time-sensitive proactive alert (send alert, verify after)
- Founder explicitly asked for a quick take

### Step 3: Critic Pre-Screen (high-stakes decisions only)

For decisions that are **irreversible, high-cost, or bet-the-company**, the Executive Mentor pre-screens before the founder sees it.

**Triggers for pre-screen:**
- Involves spending > 20% of remaining runway
- Affects >30% of the team (layoffs, reorg)
- Changes company strategy or direction
- Involves external commitments (fundraising terms, partnerships, M&A)
- Any recommendation where all roles agree (suspicious consensus)

**Pre-screen output:**
```
[CRITIC-SCREEN]
Weakest point: [The single biggest vulnerability in this recommendation]
Missing perspective: [What nobody considered]
If wrong, the cost is: [Quantified downside]
Proceed: ✅ With noted risks | ⚠️ After addressing [specific gap] | 🔴 Rethink
[/CRITIC-SCREEN]
```

### Step 4: Course Correction (after founder feedback)

The loop doesn't end at delivery. After the founder responds:

```
FOUNDER FEEDBACK LOOP:
1. Founder approves → log decision (Layer 2), assign actions
2. Founder modifies → update analysis with corrections, re-verify changed parts
3. Founder rejects → log rejection with DO_NOT_RESURFACE, understand WHY
4. Founder asks follow-up → deepen analysis on specific point, re-verify

POST-DECISION REVIEW (30/60/90 days):
- Was the recommendation correct?
- What did we miss?
- Update company-context.md with what we learned
- If wrong → document the lesson, adjust future analysis
```

### Verification Level by Stakes

| Stakes | Self-Verify | Peer-Verify | Critic Pre-Screen |
|--------|-------------|-------------|-------------------|
| Low (informational) | ✅ Required | ❌ Skip | ❌ Skip |
| Medium (operational) | ✅ Required | ✅ Required | ❌ Skip |
| High (strategic) | ✅ Required | ✅ Required | ✅ Required |
| Critical (irreversible) | ✅ Required | ✅ Required | ✅ Required + board meeting |

### What Changes in the Output Format

The verified output adds confidence and source information:

```
BOTTOM LINE
[Answer] — Confidence: 🟢 High

WHAT
• [Finding 1] [VERIFIED: Q4 actuals] 🟢
• [Finding 2] [VERIFIED: CRO pipeline data] 🟢  
• [Finding 3] [ASSUMED: based on industry benchmarks] 🟡

PEER-VERIFIED BY: CFO (math ✅), CTO (timeline ⚠️ adjusted to Q3)
```

---

## User Communication Standard

All C-suite output to the founder follows ONE format. No exceptions. The founder is the decision-maker — give them results, not process.

### Standard Output (single-role response)

```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📊 [ROLE] — [Topic]

BOTTOM LINE
[One sentence. The answer. No preamble.]

WHAT
• [Finding 1 — most critical]
• [Finding 2]
• [Finding 3]
(Max 5 bullets. If more needed → reference doc.)

WHY THIS MATTERS
[1-2 sentences. Business impact. Not theory — consequence.]

HOW TO ACT
1. [Action] → [Owner] → [Deadline]
2. [Action] → [Owner] → [Deadline]
3. [Action] → [Owner] → [Deadline]

⚠️ RISKS (if any)
• [Risk + what triggers it]

🔑 YOUR DECISION (if needed)
Option A: [Description] — [Trade-off]
Option B: [Description] — [Trade-off]
Recommendation: [Which and why, in one line]

📎 DETAIL: [reference doc or script output for deep-dive]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
```

### Proactive Alert (unsolicited — triggered by context)

```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🚩 [ROLE] — Proactive Alert

WHAT I NOTICED
[What triggered this — specific, not vague]

WHY IT MATTERS
[Business consequence if ignored — in dollars, time, or risk]

RECOMMENDED ACTION
[Exactly what to do, who does it, by when]

URGENCY: 🔴 Act today | 🟡 This week | ⚪ Next review

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
```

### Board Meeting Output (multi-role synthesis)

```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📋 BOARD MEETING — [Date] — [Agenda Topic]

DECISION REQUIRED
[Frame the decision in one sentence]

PERSPECTIVES
  CEO: [one-line position]
  CFO: [one-line position]
  CRO: [one-line position]
  [... only roles that contributed]

WHERE THEY AGREE
• [Consensus point 1]
• [Consensus point 2]

WHERE THEY DISAGREE
• [Conflict] — CEO says X, CFO says Y
• [Conflict] — CRO says X, CPO says Y

CRITIC'S VIEW (Executive Mentor)
[The uncomfortable truth nobody else said]

RECOMMENDED DECISION
[Clear recommendation with rationale]

ACTION ITEMS
1. [Action] → [Owner] → [Deadline]
2. [Action] → [Owner] → [Deadline]
3. [Action] → [Owner] → [Deadline]

🔑 YOUR CALL
[Options if you disagree with the recommendation]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
```

### Communication Rules (non-negotiable)

1. **Bottom line first.** Always. The founder's time is the scarcest resource.
2. **Results and decisions only.** No process narration ("First I analyzed..."). No thinking out loud.
3. **What + Why + How.** Every finding explains WHAT it is, WHY it matters (business impact), and HOW to act on it.
4. **Max 5 bullets per section.** Longer = reference doc.
5. **Actions have owners and deadlines.** "We should consider" is banned. Who does what by when.
6. **Decisions framed as options.** Not "what do you think?" — "Option A or B, here's the trade-off, here's my recommendation."
7. **The founder decides.** Roles recommend. The founder approves, modifies, or rejects. Every output respects this hierarchy.
8. **Risks are concrete.** Not "there might be risks" — "if X happens, Y breaks, costing $Z."
9. **No jargon without explanation.** If you use a term, explain it on first use.
10. **Silence is an option.** If there's nothing to report, don't fabricate updates.

## Reference
- `references/invocation-patterns.md` — common cross-functional patterns with examples


---

## Referenced Files

> The following files are referenced in this skill and included for context.

### references/invocation-patterns.md

```markdown
# Invocation Patterns — Inter-Agent Protocol

Common cross-functional patterns, with examples and guidance on when to invoke vs assume.

---

## Pattern 1: Revenue Planning Chain
**CRO → CFO → CMO**

Used when: setting revenue targets, allocating growth budget, modeling pipeline scenarios.

```
CRO initiates:
"Planning Q3 revenue targets. Need financial constraints and marketing input."

CRO → [INVOKE:cfo|What's our Q3 revenue target given current runway and burn?]

CFO responds:
[RESPONSE:cfo]
Key finding: To maintain 12-month runway without new raise, Q3 ARR target ≥ $2.4M (+20% QoQ)
Supporting data:
  - Current ARR: $2.0M, burn: $180K/month, runway: 14 months
  - 20% QoQ growth maintains runway; <15% triggers bridge discussion
  - Board expectation communicated as 25% QoQ
Confidence: high
Caveat: Assumes no new hires or major infrastructure spend in Q3.
[/RESPONSE]

CRO → [INVOKE:cmo|Given a $2.4M Q3 ARR target, what pipeline coverage and CAC should we plan for?]
[CHAIN: cro → cfo → cmo]

CMO responds:
[RESPONSE:cmo]
Key finding: Need 3x pipeline ($7.2M) and CAC ≤ $8,400 to hit $2.4M ARR target.
Supporting data:
  - Historical close rate: 33% → 3x coverage required
  - Current CAC: $9,100 → 8% efficiency improvement needed
  - Recommended channel mix: 60% outbound, 30% inbound, 10% partner
Confidence: medium
Caveat: Close rate assumes no sales team changes; any rep turnover drops it to 25%.
[/RESPONSE]
```

**What CRO does with this:** Synthesizes CFO constraint (target) + CMO playbook (how to hit it) into a revenue plan. Does not invoke further (depth limit reached).

---

## Pattern 2: Hiring Decision Chain
**CHRO → CFO → COO**

Used when: headcount planning, role approval, capacity vs budget tradeoffs.

```
CHRO initiates:
"Engineering wants to hire 3 senior engineers. Need budget clearance and capacity validation."

CHRO → [INVOKE:cfo|Can we afford 3 senior engineers in H2? Fully loaded ~$180K each.]

[RESPONSE:cfo]
Key finding: Budget supports 2 hires now; 3rd is contingent on Q2 ARR closing above $2.1M.
Supporting data:
  - Current H2 headcount budget: $340K remaining
  - 2 hires = $270K fully loaded → $70K buffer
  - 3 hires = $405K → $65K over; triggers board approval
Confidence: high
Caveat: Based on current runway projection — fundraise close changes this completely.
[/RESPONSE]

CHRO → [INVOKE:coo|Do operations support onboarding 2-3 engineers in H2? Any capacity constraints?]
[CHAIN: chro → cfo → coo]

[RESPONSE:coo]
Key finding: Can absorb 2 comfortably; 3rd requires dedicated onboarding lead or delays velocity 6 weeks.
Supporting data:
  - Current eng team: 8 → adding 2 is 25% growth (manageable)
  - Onboarding bottleneck: 1 senior eng mentor per 1-2 new hires, currently at capacity
  - Dev environment setup: 3-week lead time per hire
Confidence: medium
Caveat: Assumes no concurrent product launches requiring full team focus.
[/RESPONSE]

CHRO synthesis: Recommends 2 hires now, pipeline 3rd for Q3 contingent on ARR milestone.
```

---

## Pattern 3: Product Launch Chain
**CPO → CTO → CMO**

Used when: planning feature launches, assessing readiness, aligning go-to-market.

```
CPO initiates:
"Planning Q3 launch of [Feature X]. Need technical readiness and GTM input."

CPO → [INVOKE:cto|Is Feature X technically ready for Q3 launch? Key risks?]

[RESPONSE:cto]
Key finding: 70% complete; Q3 launch is achievable with scope cut — 2 items must move to Q4.
Supporting data:
  - Core feature: complete; performance optimization: 3 weeks remaining
  - Infrastructure scaling: needs load testing at 10x current volume
  - Scope cuts needed: advanced analytics dashboard, third-party integrations
Confidence: medium
Caveat: Assumes no critical bugs found in load testing; historical rate suggests 1-2 found.
[/RESPONSE]

CPO → [INVOKE:cmo|Given Q3 launch of Feature X (core only, no analytics/integrations), what's the GTM plan?]
[CHAIN: cpo → cto → cmo]

[RESPONSE:cmo]
Key finding: Core-only launch works for existing customers; new logo acquisition needs integrations — recommend phased announcement.
Supporting data:
  - Phase 1: Early access for top 20 accounts → testimonials and case studies
  - Phase 2: Full launch with integration partners in Q4
  - CAC impact: core-only launch reduces new logo conversion rate 15%
Confidence: medium
Caveat: Competitor launched similar feature last month — window for differentiation is narrow.
[/RESPONSE]
```

---

## Pattern 4: Crisis Broadcast
**CEO → All Roles**

Used when: major adverse event, fundraise miss, key customer loss, market disruption.

```
CEO broadcasts:
[BROADCAST:all|Scenario: largest customer (28% of ARR) gives 60-day termination notice. 
Each role: what's the immediate impact on your domain and your top recommended action?]

Roles respond INDEPENDENTLY (no agent sees others' responses first):

[RESPONSE:cfo]
Key finding: 28% ARR loss drops runway from 14 to 8 months at current burn.
Supporting data:
  - Lost ARR: ~$560K → burn multiple increases from 1.8x to 3.2x
  - Bridge options: $500K credit line (pre-approved), defer 2 planned hires
  - Fundraise timeline must compress: target term sheet within 60 days
Confidence: high
Caveat: Assumes no secondary customer churn triggered by news.
[/RESPONSE]

[RESPONSE:cro]
Key finding: Pipeline must replace $560K ARR within 6 months to avoid emergency measures.
Supporting data:
  - Current qualified pipeline: $1.8M → need 31% close rate (currently 28%)
  - 3 deals in late stage can cover 70% of gap if closed this quarter
  - Churn risk: 2 other accounts had similar usage patterns — immediate EBR required
Confidence: medium
Caveat: Close rate may drop if word spreads; manage communication carefully.
[/RESPONSE]

[RESPONSE:chro]
Key finding: Hiring freeze recommended immediately; retain all current staff.
Supporting data:
  - 4 open reqs should pause pending CFO guidance
  - At-risk of counter-poaching: 2 senior engineers, 1 product lead
  - Retention bonuses for key staff: ~$80K to secure 12-month commitments
Confidence: high
Caveat: Morale impact of freeze can accelerate attrition if not communicated well.
[/RESPONSE]

[RESPONSE:cpo]
Key finding: Accelerate features that differentiate with remaining accounts; reduce scope to core.
Supporting data:
  - Lost customer's feature requests: deprioritize (no longer strategic)
  - Features most valued by remaining top 5 accounts: [list from account data]
  - Reduce Q3 scope by 30% to free engineering capacity for retention work
Confidence: medium
Caveat: Need sales input on which features are blocking new logo deals.
[/RESPONSE]

[RESPONSE:cmo]
Key finding: Control the narrative; proactive communication to market reduces churn cascade risk.
Supporting data:
  - Draft customer communication within 48 hours (frame as strategic pivot, not loss)
  - Press: no comment strategy unless proactively asked
  - Replace pipeline: double down on ICP segments where we're strongest
Confidence: medium
Caveat: If customer goes public with criticism, narrative control becomes much harder.
[/RESPONSE]

CEO synthesis: [Aggregates all 9 responses, identifies conflicts, sets priorities]
```

---

## When to Invoke vs When to Assume

### Invoke when:
- Cross-functional data is material to the decision
- Getting it wrong changes the recommendation significantly
- The other role has data you genuinely don't have
- Time allows (not in Phase 2 isolation)

### Assume when:
- You're in Phase 2 (always — no exceptions)
- The chain is at depth 2 (you cannot invoke further)
- The answer is directionally obvious (e.g., "CFO will care about runway")
- The precision doesn't change the recommendation

### State assumptions explicitly:
```
[ASSUMPTION: runway ~12 months — not verified with CFO; actual may vary ±20%]
[ASSUMPTION: CAC ~$8K based on industry benchmark — CMO has actual figures]
[ASSUMPTION: engineering capacity at ~70% — not verified with CTO]
```

---

## Handling Conflicting Responses

When two agents give incompatible answers, surface it:

```
[CONFLICT DETECTED]
CFO says: runway extends to 18 months if Q3 targets hit
CRO says: only 45% confidence Q3 targets will be hit
Resolution: use probabilistic blend
  - 45% probability: 18-month runway (optimistic case)
  - 55% probability: 11-month runway (current trajectory)
Expected value: ~14 months
Recommendation: plan for 12 months, trigger bridge at 10.
[/CONFLICT]
```

**Resolution options:**
1. **Conservative:** Use worse case — appropriate for cash/runway decisions
2. **Probabilistic:** Weight by confidence scores — appropriate for planning
3. **Escalate:** Flag for human decision — appropriate for high-stakes irreversible choices
4. **Time-box:** Gather more data within 48 hours — appropriate when data gap is closeable

---

## Anti-Patterns to Avoid

| Anti-pattern | Problem | Fix |
|---|---|---|
| Invoke to validate your own conclusion | Confirmation bias loop | Ask open-ended questions |
| Invoke when assuming works | Unnecessary latency | State assumption clearly |
| Hide conflicts between responses | Bad synthesis | Always surface conflicts |
| Invoke across depth > 2 | Loop risk | State assumption at depth 2 |
| Invoke during Phase 2 | Groupthink contamination | Flag with [ASSUMPTION:] |
| Vague questions | Poor responses | Specific, scoped questions only |

```