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/checkpoint.go

@@ -16,21 +16,10 @@ package agent
 
 import (
 	"context"
-	"errors"
 
 	"go.probo.inc/probo/pkg/llm"
 )
 
-// ErrSuspendForCheckpoint is the recommended cancel cause for callers
-// who want the agent loop to gracefully suspend (build a checkpoint
-// and return *SuspendedError) rather than treat the cancellation as a
-// silent close. The agent loop only inspects ctx.Err(); any cancel
-// cause produces a graceful suspend, but using this sentinel makes
-// the intent explicit and lets supervisors distinguish a user-driven
-// cancel from infrastructure-level causes (lease loss, heartbeat
-// failure) when they inspect context.Cause(ctx).
-var ErrSuspendForCheckpoint = errors.New("agent run: graceful suspend requested")
-
 type (
 	AgentStatus string