SLOPE

Sprint Lifecycle & Operational Performance Engine

A framework for measuring and improving sprint execution quality. Replace subjective retrospectives with objective, quantifiable metrics. Works with Claude Code, Cursor, and OpenCode.

Quick Start

# Install
npm install -g @srbryers/cli

# Initialize in your project (auto-detects your AI tool)
slope init

# Or specify your platform
slope init --claude-code
slope init --cursor
slope init --opencode

# View your handicap card
slope card

# Get a pre-sprint briefing
slope briefing

# Generate an HTML performance report
slope report --html

What is SLOPE?

SLOPE maps sprint execution to a scoring metaphor. Every sprint has a par (expected ticket count), every ticket is a shot with an approach complexity and outcome. Over time, your handicap reveals patterns: Do you over-engineer? Under-scope? Pick the wrong approach?

SLOPE provides:

• Scorecards — structured sprint retros with quantified outcomes
• Handicap tracking — rolling performance windows (last 5, 10, all-time)
• Dispersion analysis — miss pattern detection and systemic issue identification
• Training recommendations — data-driven improvement suggestions
• Agent guidance hooks — real-time hints for AI coding agents
• HTML reports — self-contained visual performance dashboards
• Multi-platform support — Claude Code, Cursor, OpenCode

Metaphors

SLOPE ships with 6 built-in metaphors. The scoring math is identical — only the terminology changes.

Configure in .slope/config.json:

{ "metaphor": "gaming" }

Or per-command: slope review --metaphor=tennis

Packages

CLI Commands

Scoring & Analysis

Planning & Guidance

Sessions & Claims

Agent Hooks

Setup

Platform Setup

Claude Code

slope init --claude-code

Installs:

• .claude/rules/ — Sprint checklist, commit discipline, review loop
• .claude/hooks/ — Guard dispatcher for real-time guidance
• .mcp.json — SLOPE MCP server config
• CLAUDE.md — Project context

Cursor

slope init --cursor

Installs:

• .cursor/rules/ — SLOPE methodology rules
• .cursor/mcp.json — MCP server config

OpenCode

slope init --opencode

Installs:

• AGENTS.md — SLOPE methodology (OpenCode reads this format)
• .opencode/plugins/slope-plugin.ts — Event capture plugin

Agent Guidance Hooks

SLOPE can guide AI agents in real-time via hook integration. Six guards provide contextual hints:

Install all guards:

slope hook add --level=full

Guards are non-blocking hints (except stop-check) — they inject context, never deny actions.

Configuration

After slope init, configure .slope/config.json:

{
  "scorecardDir": "docs/retros",
  "scorecardPattern": "sprint-*.json",
  "minSprint": 1,
  "metaphor": "golf",
  "commonIssuesPath": ".slope/common-issues.json",
  "sessionsPath": ".slope/sessions.json",
  "roadmapPath": "docs/backlog/roadmap.json",
  "guidance": {
    "disabled": [],
    "commitInterval": 15,
    "pushInterval": 30
  }
}

Core API

import {
  buildScorecard,
  validateScorecard,
  computeHandicapCard,
  computeDispersion,
  formatSprintReview,
  recommendClub,
  classifyShot,
  generateTrainingPlan,
  buildReportData,
  generateHtmlReport,
  getMetaphor,
  loadScorecards,
} from '@srbryers/core';

// Build a scorecard
const card = buildScorecard({
  sprint_number: 1,
  theme: 'My First Sprint',
  par: 3,
  slope: 0,
  date: '2026-02-22',
  shots: [
    { ticket_key: 'S1-1', title: 'Setup', club: 'short_iron', result: 'green', hazards: [] },
    { ticket_key: 'S1-2', title: 'Feature', club: 'short_iron', result: 'in_the_hole', hazards: [] },
    { ticket_key: 'S1-3', title: 'Tests', club: 'wedge', result: 'green', hazards: [] },
  ],
});

// Generate HTML report with gaming metaphor
const data = buildReportData([card]);
const gaming = getMetaphor('gaming');
const html = generateHtmlReport(data, gaming);

License

MIT