Agent Ready v2.0.0

Breaking safety change: agent-ready validate is plan-only by default and no longer executes detected project commands unless --execute is passed.

Added:

• --execute is now required to run detected validation commands.
• --dry-run remains an explicit plan-only option.
• --dry-run and --execute are mutually exclusive.

Updated:

• README, skill instructions, PR package guidance, demo outputs, screenshots, and GIF now reflect the safer default.

Agent Ready v1.2.0

Added

• agent-ready.config.json scan overrides for custom commands, frameworks, entry points, important directories, generated directories, package managers, and monorepo hints.
• Regression tests for override precedence and invalid override handling.

Changed

• The generated config template now includes an explicit overrides section so complex repositories can correct heuristic misses without patching the scanner.
• README and skill docs now document override usage as the recommended answer for custom builds and complex monorepos.

Agent Ready v1.1.0

Added

• Low-noise --minimal generation mode for teams that want only AGENTS.md and core .agent-ready reports.
• AGENTS.md quality scoring for command coverage, project map, safety guidance, and maintainability.
• Strict 100-point scorecard with no raw-score cap behavior.
• Validation reports now make --dry-run safety status explicit.

Changed

• README now documents honest limits: Agent Ready is not an agent framework, not a security scanner replacement, and still uses heuristic detection.
• PR package validation guidance now recommends agent-ready validate <repo> --dry-run before executing detected commands.
• GitHub Action example now pins to hao0401/agent-ready@v1.1.0.
• Demo artifacts now normalize generated timestamps for more reproducible showcase output.

Agent Ready v1.0.0

Changelog

All notable changes to Agent Ready are documented here.

1.0.0 - 2026-06-09

Added

• Dependency-free agent-ready CLI for scanning and preparing repositories for AI coding agents.
• One-command generation for AGENTS.md, CLAUDE.md, GEMINI.md, Cursor rules, and GitHub Copilot instructions.
• Agent readiness score, scorecard, one-page summary, badge output, and full readiness report.
• Prompt-injection scan, redacted secret-looking finding scan, package-manager conflict detection, and monorepo hints.
• Baseline, baseline diff, score ratchet, dry-run validation, doctor mode, PR kit, and project-specific Codex skill generation.
• Composite GitHub Action with PR annotations, sticky PR comments, workflow summaries, outputs, and SARIF support.
• Reproducible demo repository, before/after artifacts, terminal screenshot, terminal GIF, and public showcase README.

Verified

• Release gate compiles all Python entry points, runs the regression suite, regenerates demo assets, checks local Markdown links, scans public demo outputs, verifies wheel and sdist contents, and smoke-tests the installed console command.