Add v3 doc references to CLAUDE.md, .claude/CLAUDE.md, README.md, and docs index. Add v3 planning TODOs. Update progress log with v3 session. Update CHANGELOG with v3 planning entry. Trim oldest completed TODOs.
29 lines
2.7 KiB
Markdown
29 lines
2.7 KiB
Markdown
# BTerminal -- TODO
|
|
|
|
## Active
|
|
|
|
### v3 Planning
|
|
- [ ] **Collect adversarial agent findings** -- 3 architecture agents launched (Architect, Devil's Advocate, UX+Perf Specialist). Collect and synthesize findings.
|
|
- [ ] **Produce final v3 architecture plan** -- Finalize decisions on config format, layout engine, session isolation, state management, etc.
|
|
- [ ] **Create v3 implementation phases** -- Break v3 into phased implementation plan.
|
|
|
|
### v2 Remaining
|
|
- [ ] **E2E testing (Playwright/WebDriver)** -- Scaffold at v2/tests/e2e/README.md. Needs display server to run. Test: open terminal, run command, open agent, verify output.
|
|
- [ ] **Multi-machine real-world testing** -- Test bterminal-relay with 2 machines (local + 1 remote). Verify PTY + agent operations over WebSocket.
|
|
- [ ] **Multi-machine TLS/certificate pinning** -- Add TLS support to bterminal-relay and certificate pinning in RemoteManager for production security.
|
|
- [ ] **Agent Teams real-world testing** -- Frontend routing implemented (Phase 7). Needs testing with CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 and real subagent spawning.
|
|
- [ ] **Model selector in AgentPane** -- Dropdown for model override per session (field already passes through to SDK).
|
|
- [ ] **System prompt field in AgentPane** -- Custom instructions per agent session (field already passes through to SDK).
|
|
- [ ] **Additional directories picker** -- UI for additional_directories per session (field already passes through to SDK).
|
|
|
|
## Completed
|
|
|
|
- [x] **Claude profiles & skill discovery** -- switcher-claude profile integration, skill autocomplete in agent prompt (`/` prefix), extended AgentQueryOptions with 5 new fields. | Done: 2026-03-07
|
|
- [x] **Claude CLI path auto-detection** -- findClaudeCli() in both sidecar runners auto-detects Claude CLI path, passes to SDK via pathToClaudeCodeExecutable. Early error if CLI not found. | Done: 2026-03-07
|
|
- [x] **Unified sidecar bundle** -- Single agent-runner.mjs runs on both Deno and Node.js. Rust-side CLAUDE* env var stripping (dual-layer). | Done: 2026-03-07
|
|
- [x] **AgentPane onDestroy bug fix** -- Stop-on-close moved to TilingGrid onClose handler. | Done: 2026-03-06
|
|
- [x] **Permission mode passthrough** -- permission_mode field flowing Rust -> sidecar -> SDK. | Done: 2026-03-06
|
|
- [x] **Sidecar SDK migration** -- Migrated from raw CLI spawning to @anthropic-ai/claude-agent-sdk query(). | Done: 2026-03-06
|
|
- [x] **Sidecar CLAUDE* env var leak fix** -- Strip ALL CLAUDE-prefixed env vars (dual-layer). | Done: 2026-03-06
|
|
- [x] **Multi-machine reconnection** -- Exponential backoff, TCP probe, frontend listeners + auto-reconnect. | Done: 2026-03-06
|
|
- [x] **Multi-machine support (Phases A-D)** -- bterminal-core, bterminal-relay, RemoteManager, frontend. | Done: 2026-03-06
|