AI Skills Marketplace / Sales

AI SDR Objection Handler

A branching decision skill that gives AI SDR agents a tested response tree for the objections that actually kill deals: price, timing, and authority, with clear escalation rules to a human rep.

The AI SDR Objection Handler is a free Claude Skill that routes price, timing, and authority objections through a tested 6-branch decision tree with explicit rules for when to escalate to a human sales rep instead of guessing.

↗ View Source ⬇ Download Full Repo (.zip)
MIT
License
6
Decision branches
3
Escalation rules
Yes
Multi-tool support
Aug 12, 2026
Last updated
9/9 passing
Tests

Most AI SDR objection handling is a single flat prompt: “respond to objections professionally.” This skill replaces that with a real decision tree: each objection type routes to a specific response strategy, and the tree knows when it’s out of its depth and should hand off to a human rep instead of guessing.

The 6 branches

Objection type Escalation rule
Discount ask After 2nd pushback round
Competitor price comparison After 2nd pushback round
Budget freeze After 2nd pushback round
“Not now” If stated timeline > 180 days
“Next quarter” If stated timeline > 180 days
Not the decision maker Immediately: no response attempt, request a warm intro

Multi-tool support

The same skill definition runs unmodified across Claude Code, Claude Desktop, and any MCP-compatible agent runtime. The branching logic is tool-agnostic.

Configuration

{
  "handoff_channel": "#sales-alerts",
  "max_price_pushback_rounds": 2,
  "escalate_on_authority_objection": true,
  "timing_escalation_threshold_days": 180
}

Full install steps, security & validation results, and calibration guidance are in the README below.

README.md View raw ↗

What this is (and isn’t)

This is a documentation + skill-definition repo, not a hosted service. Installing gives you the skill file and decision logic. MV3 doesn’t run this on your behalf unless you separately request implementation help.

Files

FilePurpose
SKILL.mdThe actual skill definition: load this into Claude Code/Desktop/an MCP agent
objection-tree.jsonThe 6-branch decision tree + escalation config, editable without touching code
router.jsZero-dependency decision router that reads the tree and decides respond-vs-escalate
test.jsAutomated test covering all 3 escalation rule types + a structural sanity check

Quick start

git clone https://github.com/mv3marketing/ai-skills.git
cd ai-skills/sales/ai-sdr-objection-handler
node test.js          # confirms the tree + router work before you load them

Then load SKILL.md into your agent runtime.

Install

PlatformSteps
Claude CodeCopy this directory into your project’s .claude/skills/. SKILL.md’s frontmatter is what makes it discoverable.
Claude DesktopSame skill format, same file: Desktop’s skill loading follows the same SKILL.md convention.
MCP / other agentobjection-tree.json + router.js have zero platform dependencies.

Security & validation

Verified 2026-08-12, not copied from a template: npm audit shows 0 vulnerabilities (zero runtime dependencies), node test.js passes 9/9. No outbound network calls anywhere in the decision logic.

Calibrating to your own objections

objection-tree.json’s trigger examples and response strategies reflect MV3’s own experience: a real starting point, not a universal script. Swap in your own top 5-10 recurring objections from real call transcripts before running this live.

Support

Found a bug or have a question about how this skill works? Open an issue on GitHub Issues. Want it wired into your real Slack, CRM, or dialer? That’s paid implementation work: book a scoping call.

License

MIT

FAQ

Is this skill free to install?

Yes, every skill in the marketplace is free to download and run yourself. You only pay if you want MV3 to implement or customize it for your specific stack ($175/hr).

What tools does this skill work with?

Claude Code, Claude Desktop, and any MCP-compatible agent runtime.

Does installing this skill send any of my data to MV3?

No. The skill definition runs entirely inside your own agent session with zero outbound network calls in the decision logic.

Can I modify the objection responses?

Yes, the skill ships as an editable JSON decision tree, not a locked black box.

Where do I get support if something breaks?

Open a GitHub Issue on the repo for bugs or questions. For paid implementation help wiring this into your real CRM or Slack, book a scoping call.

JR
Jordan Reeves
ABM & Outbound Pipeline, MV3 Marketing