v0.2.3

Fixes RightNow-AI#55: Cron scheduler no longer fires every minute. compute_next_run now offsets by +1 second so jobs respect their actual schedule.
Fixes RightNow-AI#100: OpenRouter model prefix stripped before API call. Models like openrouter/google/gemini-2.5-flash now send the correct ID to the provider.
Fixes RightNow-AI#83: Added OPENFANG_LISTEN env var override for Docker containers where 0.0.0.0 binding is needed.
Fixes RightNow-AI#51: Dashboard auth prompt now detects 401 correctly by checking a protected endpoint. Added /api/workflows and /api/logs/stream to the public allowlist.
Fixes RightNow-AI#65: Added 6 new OpenRouter models including free tier Llama 3.3 70B, Mistral 7B, Gemma 2 9B, Qwen 2.5 72B, and DeepSeek V3.
Fixes RightNow-AI#68: Approval policy require_approval field now accepts boolean shorthand. false clears the list, true defaults to shell_exec.
Fixes RightNow-AI#63 RightNow-AI#47 RightNow-AI#49: Custom model support via POST /api/models/custom. Persisted to disk, loaded on boot, with dashboard UI for adding models.

v0.2.2

Fixes:
- Fix cross-provider model switching (MiniMax, Gemini, all providers)
- resolve_driver() no longer passes wrong API key when switching providers
- Add provider inference from model name prefix for uncataloged models
- Add MiniMax-M2.5 to model catalog (Frontier tier)

Features:
- GitHub Copilot OAuth device flow login
- OpenAI Codex provider with credential sync
- Claude Code CLI backend driver

v0.2.1

Fixes:
- Fix cron scheduler re-firing every ~1min (RightNow-AI#55): use after() instead of upcoming() to skip current second, remove double-compute in record_success, fix UTF-8 panic in record_failure
- Fix [approval_policy] config silently ignored (RightNow-AI#68): add serde alias so both [approval] and [approval_policy] work
- Fix dashboard approve/reject buttons returning 401 (RightNow-AI#51, RightNow-AI#46): add /api/approvals/ prefix to public auth allowlist

bump v0.2.0

v0.1.9 — Bug Fixes

Fixes:
- RightNow-AI#55: Cron scheduler no longer fires every minute — pre-advances next_run before execution, uses Skip tick behavior
- RightNow-AI#51: Dashboard 401s resolved — auth detection fixed, 15 dashboard endpoints added to whitelist
- RightNow-AI#46: Approval approve/reject buttons now render correctly — switched from x-show to x-if template
- RightNow-AI#44: CLI no longer hangs on --version/status — normalizes 0.0.0.0 to 127.0.0.1, adds connect_timeout

v0.1.8 — Real Email & channel_send Tool

Features:
- Real SMTP sending via lettre (STARTTLS/implicit TLS, credentials auth)
- Real IMAP polling via imap crate (TLS, UNSEEN search, mailparse RFC822)
- Email threading via In-Reply-To headers (DashMap reply context)
- New channel_send tool for agents to proactively send messages on any channel
- KernelHandle.send_channel_message() wired through kernel to channel adapters
- Channel adapters registered in kernel DashMap at bridge startup

bump version

MCP guidance

v0.1.5

Fixes:
- Fix RightNow-AI#38: panic on Chinese/multi-byte text in cap_str (byte index not char boundary)
- Fix RightNow-AI#35: improved error message for LLM failures (was "Invalid request format", now "Check your API key and model configuration")
- Fix 13 UTF-8 boundary panics across compactor, session_repair, loop_guard, image_gen, tool_runner, tts, workspace_context
- Added safe_truncate_str utility for safe multi-byte string truncation

v0.1.4

- Editable provider base URLs for local models (Ollama/vLLM/LM Studio)
- PUT /api/providers/{name}/url endpoint with validation + health probe
- provider_urls config section persisted to config.toml
- Hot-reload support for provider URL changes
- Dashboard UI: base URL input for local providers
- base_url field added to GET /api/providers response