What Tablif read on agent-roast's site
agent-roast describes itself as: “GitHub - Nari-Ab/agent-roast: Zero-telemetry CLI scorecard auditing your git history for AI coding agent infractions, shortcuts, and panic loops. · GitHub”
- Stage: Live, no waitlist
What it is
Calls itself “CLI scorecard”.
- Calls itself: CLI scorecard
“Zero-telemetry CLI scorecard auditing your git history for AI coding shortcuts, fake tests, and panic loops.”
Where it puts itself
Claims “100%”. Rules out “pull_request_target”.
- Superlative claim: 100%
- Says it will not: pull_request_target
“100% offline, zero API keys.”
“…pull_request_target is not supported.”
Who it sells to
Sells to developers.
- Sells to: developers
“Zero-telemetry CLI scorecard auditing your git history for AI coding shortcuts, fake tests, and panic loops.”
How it runs
A command-line tool. Runs locally and works offline. Distributed through the npm. No API. Open source. AI is a side feature.
- Kind: cli
- Runs where: on the device, offline
- Distributed via: npm
- API: no
- Open source: yes
- AI role: a side feature
“Zero-telemetry CLI scorecard auditing your git history for AI coding shortcuts, fake tests, and panic loops.”
“100% offline, zero API keys.”
What it works with
Analyzes / reviews Aider, Antigravity, Claude Code and 2 more. Names GitHub Copilot on its pages.
- Analyzes / reviews: Aider · Antigravity · Claude Code · +2
- Names: GitHub Copilot
“…agent-roast scans recorded git diffs for shortcuts taken by AI coding assistants (Claude Code, Cursor, Codex, Antigravity, Aider).”
“GitHub Copilot Write better code with AI”
What it offers as proof
Live, no waitlist. Claims “304 AI commits”. Publishes docs.
- Number claimed: 304 AI commits
“…agent-roast — git discipline audit (304 AI commits in Soup)”
“…docs -> https://github.com/Nari-Ab/agent-roast/tree/main/docs”
Sources
Tablif read these pages, last on 16 Oct 2026. Everything above is what the product says about itself; Tablif does not verify every claim.
Similar products in Tablif Town
- Git Repository Agent: AI-powered GitHub repository analysis
- TheGitAI: Agentic AI coding tool for your terminal
- The Agentic Coding Kit: Rule files that make AI coding agents ship trustworthy code
- GitViber: Run coding agents in parallel worktrees, review diffs live
- DebugBundle: Production debugging for AI coding agents
- Terminalblog.com: Tracking the AI coding agent ecosystem
- Agent-watchdog: Runtime loop detection for LangGraph agents
- AI Visibility Checker: Open-source audit for SEO, AI search and AI agents