Tighten ctx-suspend plumbing and trim docs

Address review feedback:

- Move ErrSuspendForCheckpoint from checkpoint.go to errors.go
  next to the rest of the agent error declarations; drop the
  colon in the error string so it matches the existing
  `agent run <event>` style used by the supervisor sentinels.
- Replace the inline `outerCtx := ctx; ctx = context.WithoutCancel(ctx)`
  pattern with a small `suspendShield` helper in context.go used
  by coreLoop, resumeWithOpts, and resumeNested. Reads more
  cleanly and stops surfacing the WithoutCancel mechanism at
  every call site.
- Trim the doc comments on Run, RunStreamed, Resume, Restore, the
  ErrSuspendForCheckpoint declaration, and the saveCtx comment in
  restoreNestedSuspended down to the contract bullet.

Signed-off-by: Aurélien Sibiril <81782+aureliensibiril@users.noreply.github.com>

pkg/agent/stream.go

@@ -61,9 +61,7 @@ func (sr *StreamedRun) Wait() (*Result, error) {
 
 // RunStreamed launches the agent loop and returns immediately with a
 // StreamedRun whose Events channel emits incremental progress. ctx
-// follows the same graceful-suspend contract as Run: cancellation
-// triggers a checkpoint at the next safe boundary and the run finishes
-// with a *SuspendedError surfaced via Wait. See Run for details.
+// follows Run's graceful-suspend contract.
 func (a *Agent) RunStreamed(ctx context.Context, messages []llm.Message, opts ...RunOption) *StreamedRun {
 	events := make(chan StreamEvent, 64)
 	sr := &StreamedRun{