Expand digest response coverage from paired outcome evidence

[devin-ai-integration]

Goal

Expand compact/digest response coverage only where paired output/outcome measurements identify a
real token sink and prove that actionability is preserved.

Scope

After the routine workflow measurement issue lands:

1. rank command surfaces by combined CLI/MCP bytes and follow-up call cost;
2. select one high-volume surface without an adequate digest;
3. add a conservative ResponseView that retains stable identity, actionable refs/handles,
   warnings, recovery fields, and artifacts;
4. compare bytes, next-command success, fallback observations, and retries;
5. keep compatibility-sensitive defaults unchanged unless separately versioned.

Constraints

• Do not add a blanket compact default.
• Do not treat byte reduction as success when follow-up calls increase.
• Do not gate on live simulator output sizes.
• One response family per PR.

Blocked by: the paired routine output/actionability/recovery benchmark.

Origin: post-#1174 architecture follow-up.

[devin-ai-integration]

Completed for the selected response family by #1208 and finalized in #1213: network digest is 26,371 → 1,665 bytes while retaining all entries and recovery/actionability data. Any further family requires a new paired byte/outcome case.