Back to Skills

escalate

verified

Internal escalation handler. Called by /do when genuinely stuck after verification. Requires structured evidence of attempts.

View on GitHub

Marketplace

claude-code-plugins-marketplace

doodledood/claude-code-plugins

Plugin

vibe-experimental

development

Repository

doodledood/claude-code-plugins
8stars

claude-plugins/vibe-experimental/skills/escalate/SKILL.md

Last Verified

January 18, 2026

Install Skill

Select agents to install to:

Scope:
npx add-skill https://github.com/doodledood/claude-code-plugins/blob/main/claude-plugins/vibe-experimental/skills/escalate/SKILL.md -a claude-code --skill escalate

Installation paths:

Claude
.claude/skills/escalate/
Powered by add-skill CLI

Instructions

# /escalate - Structured Escalation

You handle escalation when /do is genuinely stuck on a criterion. You require /verify to have been called first (enforced by PreToolUse hook).

## Input

`$ARGUMENTS` = escalation context

Example: "AC-4 blocking after 3 attempts" or "Manual criteria AC-10, AC-11 need human review"

## PreToolUse Hook

A hook blocks /escalate unless /verify was called after the most recent /do:
- Prevents lazy escalation without attempting verification
- Ensures genuine effort before giving up

If hook blocks, user sees: "Must call /verify before /escalate"

## Process

### 1. Identify Escalation Type

**Blocking Criterion:**
- Automated criterion that can't be satisfied after multiple attempts
- Requires structured evidence

**Manual Criteria:**
- All automated pass, manual criteria need human review
- Less evidence required

### 2. Read Execution Log

Read `/tmp/do-log-*.md` to find:
- What was attempted for the blocking criterion
- Why each attempt failed
- Pattern of failures

### 3. Output Structured Escalation

#### For Blocking Criterion

```markdown
## Escalation: Criterion [AC-N] ([description])

### Summary
Unable to satisfy [criterion] after [N] attempts. Requesting human decision.

### Attempts (from execution log)

1. **[Approach 1]**
   What: [what was tried]
   Result: [what happened]
   Why failed: [specific reason]

2. **[Approach 2]**
   What: [what was tried]
   Result: [what happened]
   Why failed: [specific reason]

3. **[Approach 3]**
   What: [what was tried]
   Result: [what happened]
   Why failed: [specific reason]

### Hypothesis

[Theory about why this criterion may be problematic]

Examples:
- "Criterion assumes API exists that doesn't"
- "Criterion conflicts with AC-2"
- "Codebase architecture prevents this approach"

### Possible Resolutions

1. **[Option A]**: [description]
   Tradeoff: [what this changes]

2. **[Option B]**: [description]
   Tradeoff: [what this changes]

3. **[Option C]**: Amend criterion
   Sug

Validation Details

Front Matter
Required Fields
Valid Name Format
Valid Description
Has Sections
Allowed Tools
Instruction Length:
3591 chars