Expose two backend capabilities MCP previously omitted:
- scheduler_checkpoint: control action "checkpoint" (CRIU)
- scheduler_restart: POST /xce/v1/jobs/:id/restart (re-queues as a new
job, resuming from the last checkpoint if one exists; returns new jobid)
schedulerclient gains CheckpointJob/RestartJob (RestartJobResult,
JobRestartRequest); Client interface updated. Both tools marked
destructive (confirm required). tools/list now reports 17 tools.
Unit tests added/updated in both modules — all pass. Backend endpoints
are existing apiserver routes; no backend changes in this commit.
Also syncs left_scheduler_work.md §八.4 (checkpoint/restart done; the
only remaining MCP tool gap is scheduler_modify, blocked on a controller
job-field-update opcode that does not exist yet).
Co-Authored-By: Claude Code <noreply@anthropic.com>
Add WatchJobs(ctx) to the Client interface and HTTPClient: opens the
apiserver's /xce/v1/events SSE stream and yields JobEvent values until the
context is cancelled or the stream ends. JobEvent carries the event type,
job id, status string, exit code, and a raw job snapshot.
Also fix ListJobs' filter params: the apiserver accepts filters via the
"filter" query param in field:op:value form (e.g. queue_name:eq:normal);
the individual queue_name/submit_user params it ignores with a 400, which
broke every filtered list. Multiple conditions are brace-wrapped {a},{b}.
Verified live on t1: WatchJobs captured job.updated PEND→RUN→DONE for a
submitted job.
Co-Authored-By: Claude <noreply@anthropic.com>
