TabyAgent

v1.0.0 Breaking

This release includes breaking changes for platform teams planning a safe upgrade.

Published 15d AI Agents & Assistants

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent agentic-ai ai ai-agent ai-agents hermes

+4 more

hermes-agent openclaw telegram telegrambot

Affected surfaces

auth breaking_upgrade deps

ReleasePort's take

Light signal

editorial:auto 11d

ReleasePort Layer 1 adds several new features and bug fixes across core functionality, authentication, UI, moderation, skill handling, browser automation, container tools, and task processing.

Why it matters: New skills‑based self‑improvement, Codex OAuth, responsive quizzes, NSFW settings, expanded file‑type support, X11 controls, and critical bug fixes for ethical‑issue labeling affect developers, SREs, and security engineers managing these components.

Summary

AI summary

Updates feat, EN, and KO across a mixed release.

Changes in this release

Type	Severity	Summary	CVE
Feature
Feature	Medium	Adds Skills-based automatic self-improvement feature Adds Skills-based automatic self-improvement feature Source: llm_adapter@2026-07-15 Confidence: high	—
Feature	Medium	Adds Codex OAuth provider capability Adds Codex OAuth provider capability Source: llm_adapter@2026-07-15 Confidence: high	—
Feature	Medium	Adds visualization and responsive quiz feature Adds visualization and responsive quiz feature Source: llm_adapter@2026-07-15 Confidence: high	—
Feature	Medium	Adds NSFW speech restriction settings (block/allow indirect/allow) Adds NSFW speech restriction settings (block/allow indirect/allow) Source: llm_adapter@2026-07-15 Confidence: high	—
Feature	Medium	Adds support for docx, pdf, pptx, xlsx, frontend-design, and skill-creator skills Adds support for docx, pdf, pptx, xlsx, frontend-design, and skill-creator skills Source: llm_adapter@2026-07-15 Confidence: high	—
Feature	Medium	Adds X11 app control tools (Docker‑only) Adds X11 app control tools (Docker‑only) Source: llm_adapter@2026-07-15 Confidence: high	—
Dependency	Medium	Updates browser-use version and removes playwright dependency Updates browser-use version and removes playwright dependency Source: llm_adapter@2026-07-15 Confidence: high	—
Bugfix	Medium	Fixes model refusing tasks by labeling them "ethical issues" Fixes model refusing tasks by labeling them "ethical issues" Source: llm_adapter@2026-07-15 Confidence: high	—
Bugfix	Medium	Fixes skill‑author skills containing incorrect instructions Fixes skill‑author skills containing incorrect instructions Source: llm_adapter@2026-07-15 Confidence: high	—

Full changelog

Full Changelog: https://github.com/gpdir16/tabyAgent/compare/0.7.0...1.0.0

EN

feat: Added Skills-based automatic self-improvement (Similar to Hermes but less aggressive)
feat: Added Codex OAuth provider
feat: Added visualization and responsive quiz
feat: Added NSFW speech restriction settings (Block, allow only indirect mentions, allow)
feat: Added docx, pdf, pptx, xlsx, frontend-design, skill-creator skills
feat: Updated browser-use version (Also, removed playwright)
feat: Added X11 app control tools (Only available in Docker environment)
fix: Fixed the issue where the model was refusing tasks by calling them "ethical issues"
fix: Fixed the issue where skill-author skills were containing incorrect instructions