Add agenttrace — local TUI for AI coding-agent observability (#18)

53★ MIT local terminal UI (luoyuctl) for auditing AI coding-agent runs: token usage, costs, tool failures, latency, anomalies, diffs, and CI gates. Added to CLI & Terminal Tools. Rebased onto main to resolve same-line conflict with #27.

Author: luoyuctl

README.md

@@ -408,6 +408,7 @@ Tools for specific coding workflows.
 | [Claude Task Master](https://github.com/eyaltoledano/claude-task-master) | Free | Project management | AI-driven task management for dev projects. Integrates with Claude. |
 | [vibe-kanban](https://github.com/BloopAI/vibe-kanban) | Free | Agent orchestration | Kanban-style control plane for coordinating AI coding agents. |
 | [Ivy Tendril](https://github.com/Ivy-Interactive/Ivy-Tendril) | Free | Agent orchestration | Open-source AI coding orchestrator. Manages Claude Code, Codex, Antigravity through plan-based lifecycle with verification gates and self-improving memory. |
+| [agenttrace](https://github.com/luoyuctl/agenttrace) | Free | Session observability | Local TUI for auditing AI coding-agent runs, token usage, costs, tool failures, latency, anomalies, diffs, and CI gates. |
 | [GPT Pilot](https://github.com/Pythagora-io/gpt-pilot) | Free | Full app dev | Builds entire apps from scratch. Interactive development with AI. |
 | [Sweep](https://sweep.dev/) | Free tier | GitHub PRs | AI junior developer. Handles issues and creates PRs automatically. |