FoneClaw v0.0.7

What's new in FoneClaw 0.0.7

• Added a Top scenario acceptance list and a structured failure sample library for release regression, diagnosis, and follow-up tracking.
• Improved Agent result presentation so model errors, provider failures, HTML error pages, internal JSON, and protocol fragments are not shown directly to users.
• Improved approval cards and tool execution state tracking, including clearer accepted, rejected, cancelled, interrupted, completed, and unknown-outcome states.
• Added task-list based Agent execution guards to make multi-step tasks less repetitive and easier to finalize.
• Improved camera, screenshot, cross-app page reading, calendar creation, feedback submission, image summary, and model configuration reliability.
• Added the foundation for local L1 personal memory, including memory types, candidate extraction, policy filtering, Room storage, prompt injection, migration, and a User Profile settings page.
• Added durable Smart Memory outbox and receipt handling so background memory extraction can recover across process restarts without duplicating raw user text.
• Improved privacy protection with sensitive data redaction, safer diagnostic log formatting, backup exclusions for memory and agent data, and HTTP privacy tests.

Artifacts

• app-onlineTts-release.apk: 78,208,936 bytes, SHA-256 be2632b17862584230c6c676b26a14b2ca3a0855883363266d8f31cd47cdd4c1
• app-offlineTts-release.apk: 126,045,983 bytes, SHA-256 f0f0334c897f06fc474d9fea5516592c86da889bf4753e971a0dd9edaeaa1b60