explore-agents

name	explore-agents
description	Browse, understand, and run your saved Octave agents. Use when user says "show my agents", "list agents", "run an agent", "which agents do I have", or asks about saved agent configurations.

/octave:explore-agents - Agent Manager

Browse your saved Octave agents, understand what each does, and run them directly from Claude Code. Agents are pre-configured AI workflows for email sequences, content generation, call prep, enrichment, and qualification.

Usage

/octave:explore-agents [subcommand] [options]

Subcommands

/octave:explore-agents                      # List all agents
/octave:explore-agents list                 # List all agents (same as above)
/octave:explore-agents list --type email    # List email agents only
/octave:explore-agents show <name>          # Show agent details
/octave:explore-agents run <name> [inputs]  # Run an agent
/octave:explore-agents suggest              # Suggest agents for your task

Examples

/octave:explore-agents                                           # See all agents
/octave:explore-agents list --type email                         # Email sequence agents
/octave:explore-agents show "Enterprise Cold Outreach"           # Agent details
/octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail]
/octave:explore-agents suggest "I need to write a cold email"    # Get agent recommendations

Instructions

When the user runs /octave:explore-agents:

Subcommand: list (default)

Show all available agents:

list_agents({
  limit: 50
})

Output:

See list-output.md for the list output template.

Subcommand: list --type

Filter by agent type:

list_agents({
  type: "EMAIL"  // or "CONTENT", "CALL_PREP", "ENRICH_PERSON", "ENRICH_COMPANY", "QUALIFY_PERSON", "QUALIFY_COMPANY"
})

Types:

• email - Email sequence agents
• content - Content generation agents
• call-prep - Call preparation agents
• enrich - Enrichment agents (person + company)
• qualify - Qualification agents (person + company)

Subcommand: show

Show detailed information about an agent:

Agent: Enterprise Cold Outreach
===============================
oId: ca_abc123
Type: EMAIL
Created: 2025-12-01
Last Updated: 2026-01-15

Description:
Cold outreach sequence for enterprise prospects. Focuses on ROI,
executive pain points, and enterprise-grade proof points. Uses
formal professional tone.

Configuration:
- Sequence Length: 4 emails
- Playbook: Enterprise Sales (pb_xyz789)
- Persona Target: CTO, VP Engineering
- Brand Voice: Enterprise Professional
- Writing Style: Consultative

Email Flow:
1. Initial outreach - Pain point + social proof
2. Value follow-up - ROI story + case study
3. Insight share - Industry trend + offer
4. Breakup - Direct ask + value recap

Best For:
- Enterprise prospects ($1M+ revenue)
- Technical decision makers
- Cold outreach (no prior relationship)

---

Run this agent:
/octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail]

Or with context:
/octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail] --context "Met at conference"

Subcommand: run

Run an agent with the specified inputs:

For Email Agents:

/octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail]
/octave:explore-agents run "Enterprise Cold Outreach" --to "John Smith" --company acme.com
/octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail] --context "Referral from Sarah"

Maps to:

run_email_agent({
  agent: "Enterprise Cold Outreach",
  person: {
    email: "[e-mail]",
    // or
    firstName: "John",
    lastName: "Smith",
    companyDomain: "acme.com"
  },
  allEmailsContext: "Referral from Sarah"  // if --context provided
})

For Content Agents:

/octave:explore-agents run "LinkedIn Message Writer" --to [e-mail]
/octave:explore-agents run "Executive Brief Generator" --company acme.com

Maps to:

run_content_agent({
  agent: "LinkedIn Message Writer",
  person: { email: "[e-mail]" }
})

run_content_agent({
  agent: "Executive Brief Generator",
  company: { domain: "acme.com" }
})

For Call Prep Agents:

/octave:explore-agents run "Discovery Call Prep" --for [e-mail]
/octave:explore-agents run "Discovery Call Prep" --for [e-mail] --context "Follow-up from demo"

Maps to:

run_call_prep_agent({
  agent: "Discovery Call Prep",
  person: { email: "[e-mail]" },
  meetingContext: "Follow-up from demo"
})

For Enrichment Agents:

/octave:explore-agents run "Person Deep Dive" --person [e-mail]
/octave:explore-agents run "Company Intelligence" --company acme.com

Maps to:

run_enrich_person_agent({
  agent: "Person Deep Dive",
  person: { email: "[e-mail]" }
})

run_enrich_company_agent({
  agent: "Company Intelligence",
  company: { domain: "acme.com" }
})

For Qualification Agents:

/octave:explore-agents run "ICP Scorer" --person [e-mail]
/octave:explore-agents run "Account Fit Check" --company acme.com

Maps to:

run_qualify_person_agent({
  agent: "ICP Scorer",
  person: { email: "[e-mail]" }
})

run_qualify_company_agent({
  agent: "Account Fit Check",
  company: { domain: "acme.com" }
})

Output Format:

After running an agent, present the output clearly:

See run-output.md for the run output template.

Subcommand: suggest

Recommend agents based on task description:

/octave:explore-agents suggest "I need to write a cold email to a CTO"

Process:

1. Parse the task description
2. Match to relevant agent types
3. Return ranked recommendations

Output:

AGENT RECOMMENDATIONS
=====================
Task: "Write a cold email to a CTO"

Best Matches:
-------------

1. Enterprise Cold Outreach (95% match)
   Type: Email
   Why: Designed for technical decision makers, CTO persona targeting

   Run: /octave:explore-agents run "Enterprise Cold Outreach" --to <cto email>

2. SMB Quick Touch (70% match)
   Type: Email
   Why: Good for email sequences, but less CTO-focused

   Run: /octave:explore-agents run "SMB Quick Touch" --to <email>

---

Alternative Approach:
If you need more control, use /octave:generate email instead.
This generates emails without a pre-configured agent template.

---

Would you like me to run one of these agents?

Smart Agent Selection

When user describes a task without using /octave:explore-agents, consider suggesting relevant agents:

User: "I need to send a cold email to the CTO at acme.com"

You: I can help with that! You have an agent that's perfect for this:

Enterprise Cold Outreach - designed for technical decision makers

Would you like me to:
1. Run the "Enterprise Cold Outreach" agent for this CTO
2. Generate a custom email instead (more control, less consistency)
3. Show me other email agents first

Your choice:

Agent Types Reference

Type	MCP Tool	Purpose
EMAIL	run_email_agent	Multi-step email sequences
CONTENT	run_content_agent	General content (LinkedIn, briefs)
CALL_PREP	run_call_prep_agent	Meeting preparation
ENRICH_PERSON	run_enrich_person_agent	Person research
ENRICH_COMPANY	run_enrich_company_agent	Company research
QUALIFY_PERSON	run_qualify_person_agent	Person ICP scoring
QUALIFY_COMPANY	run_qualify_company_agent	Company ICP scoring

MCP Tools Used

Agent Management

• list_agents - List all agents with optional type filter

Agent Execution

• run_email_agent - Run email sequence agent
• run_content_agent - Run content generation agent
• run_call_prep_agent - Run call prep agent
• run_enrich_person_agent - Run person enrichment agent
• run_enrich_company_agent - Run company enrichment agent
• run_qualify_person_agent - Run person qualification agent
• run_qualify_company_agent - Run company qualification agent

Error Handling

Agent Not Found:

Agent "XYZ" not found.

Available agents:

• Enterprise Cold Outreach
• SMB Quick Touch
• [list others...]

Use /octave:explore-agents list to see all agents.

Missing Required Input:

Agent "Enterprise Cold Outreach" requires a person to target.

Usage: /octave:explore-agents run "Enterprise Cold Outreach" --to

Examples:

• /octave:explore-agents run "Enterprise Cold Outreach" --to [e-mail]
• /octave:explore-agents run "Enterprise Cold Outreach" --to "John Smith" --company acme.com

No Agents Configured:

You don't have any agents configured yet.

Agents are created in the Octave web app:

1. Go to Agents in Octave
2. Create a new agent
3. Configure its settings

Once created, you can run them from here.

In the meantime, use /octave:generate for one-off content generation.

Related Skills

• /octave:generate - Generate content with 3-way mode choice (saved agent, Octave default, Claude direct). Its --mode agent routes through agent discovery and execution — same agents you manage here.
• /octave:research - Research before running agents
• /octave:pmm - Create content that could become agent templates
• /octave:qual-doctor - Tune qualification agents (a specific agent type) with test-driven diagnosis of their scoring behavior