docs: 📚 update documentation for enhanced claude thinking sanitization and remove obsolete todo file

fix(antigravity): 🐛 improve function call response pairing with recovery strategies

refactor(antigravity): 🔨 migrate thinking sanitization to gemini message format

refactor(antigravity-claude): refactor the claude sanitization logic to prevent errors on compaction, model switching, and allow thinking.

fix(provider): 🐛 ensure claude-opus-4-5 always uses thinking variant

Merge branch 'Antigravity' of https://github.com/Mirrowel/LLM-API-Key-Proxy into fix/quota-exceeded-global-cooldown

feat(provider): add support for Claude Opus 4.5 model

fix(rotator): prevent quota errors from global cooldown

fix(logging): preserve full credential filenames in logs

fix(provider): 🐛 increase timeout for antigravity API requests from 120s to 600s

Merge pull request #14 from Mirrowel/Error-handling-consistency

fix: improve error handling implementation based on code review

feat(provider): ✨ implement Google Cloud onboarding flow with automatic project discovery

refactor(provider): 🔨 replace hardcoded project generation with dynamic GCP resolution

feat(core): ✨ add structured error accumulator and consistent error handling/reporting

fix(provider): 🐛 add automatic ID repair for mismatched tool call responses

refactor(ui): replace console.clear with cross-platform clear_screen for legacy terminal support

refactor(ui): 🔨 replace console.clear with cross-platform clear_screen function

refactor(api): 🔨 change is_ready from method to property access

fix(provider): 🐛 skip file operations for env:// credential paths

refactor(logging): 🔨 remove debug print statements and add concise deployment logs

debugging pass to try to unfuck deployment

fix(provider): 🐛 expand JSON schema validation keyword filtering and improve Gemini 3 tool call reliability

feat(provider): ✨ add strict JSON schema enforcement for Gemini 3 tool calls

fix(provider): 🐛 improve Gemini 3 tool schema handling and parameter validation

feat(api): ✨ add model pricing and capabilities enrichment service

feat(credentials): ✨ add support for environment-based credential loading and bulk export tools

chore(config): 🧹 ignore environment files and increase default token limit

feat(settings): ✨ add provider-specific settings management UI

fix(proxy): 🐛 load environment variables before displaying PROXY_API_KEY

fix(providers): 🐛 improve finish_reason handling and tool_calls initialization in stream reassembly

fix(auth): 🐛 add exponential backoff and validation for token refresh failures

feat(auth): ✨ add environment variable-based OAuth credential support with multi-account capability

feat(antigravity): ✨ add default safety settings to prevent content filtering

chore(antigravity): 🧹 update User-Agent header to version 1.11.9

feat(antigravity): ✨ add thinking mode sanitization for Claude API compatibility

fix(gemini): 🐛 consolidate parallel tool responses and improve rate limit handling

One huge ass bugfix i can't even list here. It's a mess i'll fix later

refactor(credential-tool): 🔨 add export submenu for credential management

docs: 📚 add comprehensive documentation for new features and providers

feat(rotation): ✨ add configurable weighted random credential selection

feat(client): ✨ add credential prioritization system for tier-based model access

refactor(antigravity): 🔨 remove thinking mode toggling feature

fix(client): 🐛 prevent provider initialization without configured credentials

fix(antigravity): 🐛 handle malformed double-encoded JSON responses

fix(proxy): 🐛 prevent role field concatenation in streaming responses

feat(antigravity): ✨ add thinking mode toggling for mid-conversation model switches

refactor(logging): 🔨 adjust logging levels and improve schema cleaning for Antigravity

refactor(gemini): 🔨 remove redundant model and project fields from request payload

feat(gemini): ✨ implement Gemini 3 support with tool fixes and signature caching