mcp 2.0.0 removed mcp.server.fastmcp; MCPServer is the same object under a new name, so @mcp.tool() and mcp.run() are untouched and every tool keeps its name, signature and returned shape. Verified against a clean 2.0.0 venv: all four tools list, and the whole suite -- including test_mcp.py, which could not even import while the dependency was broken -- passes at 100% coverage, 901 tests. Two things the port buys, neither of which existed in 1.x: * Tool annotations. The section banner here already said "state-only; never expose secrets, never mutate"; now the protocol says it too, as read-only, idempotent and closed-world hints, which is what lets a client run these without asking first. The day one of them starts writing, the annotation is what has to change with it. * Structured output. 2.x derives an output schema from the return types and sends structured content alongside the text, so a caller no longer re-parses JSON we already had. No code change -- it falls out of the type hints that were already there. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BsXnTHyuuCUZsqetTaajcF |
||
|---|---|---|
| .github/workflows | ||
| docs | ||
| screen_locker | ||
| scripts | ||
| stronglift_replacement | ||
| .gitignore | ||
| .mcp.json | ||
| .pre-commit-config.yaml | ||
| CLAUDE.md | ||
| DESIGN_AUDIT_TODO.md | ||
| early-bird-workout-check.timer | ||
| install_autostart.sh | ||
| install_systemd.sh | ||
| install.sh | ||
| pyproject.toml | ||
| README.md | ||
| remove_autostart.sh | ||
| remove_systemd.sh | ||
| requirements.txt | ||
| run.sh | ||
| workout-locker.service | ||
screen-locker
Tkinter/systemd screen locker with workout tracking, sick-day management, and wake-alarm integration. It enforces a workout cadence by locking the screen until a RunnerUp-verified (or explicitly justified) workout is logged.
MCP server (Claude Code integration)
screen-locker exposes a read-only MCP server (screen_locker._mcp) so
Claude Code and its subagents can query workout compliance and the
lock-decision state through typed tools — without shelling out to the
screen-locker-status CLI or opening the Tk window.
- Read-only tools:
get_status(full status snapshot),get_summary(one-line i3blocks summary +ok/warn/lockstate word),explain_lock(why the screen is / isn't locked right now),get_flags(the individual boolean lock-decision predicates for today). - There are no write/action tools, by design. Nothing here logs a workout or mutates state — workouts are logged only from RunnerUp-verified TCX data, never from a caller's claim. No tool exposes the sync token or any HMAC key.
The mcp SDK is an optional dependency (pip install -e '.[mcp]'), kept out of
the CLI/systemd system-python path. One-time setup of the dedicated venv that
Claude Code spawns:
./scripts/setup_mcp.sh
Registration lives in the checked-in .mcp.json (project scope).
Restart Claude Code in this repo and approve the project MCP-server prompt for it
to load. Verify with claude mcp list.