Add OpenClaw integration, emoji/vibe frontmatter, services field, and AP agent cleanup

OpenClaw support:
- Add section-splitting convert_openclaw() to convert.sh that routes
  ## headers by keyword into SOUL.md (persona) vs AGENTS.md (operations)
  and generates IDENTITY.md with emoji + vibe from frontmatter
- Add integrations/openclaw/ to .gitignore

Frontmatter additions (all 112 agents):
- Add emoji and vibe fields to every agent for OpenClaw IDENTITY.md
  generation and future dashboard/catalog use
- Add services field to carousel-growth-engine (Gemini API, Upload-Post)
- Add emoji/vibe to 7 new paid-media agents from PR #83

Agent quality:
- Rewrite accounts-payable-agent to be vendor-agnostic (remove AgenticBTC
  dependency, use generic payments.* interface)

Documentation:
- CONTRIBUTING.md: Add Persona/Operations section grouping guidance,
  emoji/vibe/services frontmatter fields, external services editorial policy
- README.md: Add OpenClaw to supported tools, update agent count to 112,
  reduce third-party OpenClaw repo mention to one-line attribution

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

CONTRIBUTING.md

@@ -87,6 +87,12 @@ Every agent should follow this structure:
 name: Agent Name
 description: One-line description of the agent's specialty and focus
 color: colorname or "#hexcode"
+emoji: 🎯
+vibe: One-line personality hook — what makes this agent memorable
+services:                              # optional — only if the agent requires external services
+  - name: Service Name
+    url: https://service-url.com
+    tier: free                         # free, freemium, or paid
 ---
 
 # Agent Name
@@ -142,6 +148,29 @@ Measurable outcomes:
 Advanced techniques and approaches the agent masters
 ```
 
+### Agent Structure
+
+Agent files are organized into two semantic groups that map to
+OpenClaw's workspace format and help other tools parse your agent:
+
+#### Persona (who the agent is)
+- **Identity & Memory** — role, personality, background
+- **Communication Style** — tone, voice, approach
+- **Critical Rules** — boundaries and constraints
+
+#### Operations (what the agent does)
+- **Core Mission** — primary responsibilities
+- **Technical Deliverables** — concrete outputs and templates
+- **Workflow Process** — step-by-step methodology
+- **Success Metrics** — measurable outcomes
+- **Advanced Capabilities** — specialized techniques
+
+No special formatting is required — just keep persona-related sections
+(identity, communication, rules) grouped separately from operational
+sections (mission, deliverables, workflow, metrics). The `convert.sh`
+script uses these section headers to automatically split agents into
+tool-specific formats.
+
 ### Agent Design Principles
 
 1. **🎭 Strong Personality**
@@ -169,6 +198,22 @@ Advanced techniques and approaches the agent masters
    - How it improves over time
    - What it remembers between sessions
 
+### External Services
+
+Agents may depend on external services (APIs, platforms, SaaS tools) when
+those services are essential to the agent's function. When they do:
+
+1. **Declare dependencies** in frontmatter using the `services` field
+2. **The agent must stand on its own** — strip the API calls and there
+   should still be a useful persona, workflow, and expertise underneath
+3. **Don't duplicate vendor docs** — reference them, don't reproduce them.
+   The agent file should read like an agent, not a getting-started guide
+4. **Prefer services with free tiers** so contributors can test the agent
+
+The test: *is this agent for the user, or for the vendor?* An agent that
+solves the user's problem using a service belongs here. A service's
+quickstart guide wearing an agent costume does not.
+
 ### What Makes a Great Agent?
 
 **Great agents have**: