[output]
max_summary_tokens = 500
narrative_schema = """
NOTES: <one-line what you changed>
PR: <url or none>
NEXT: <optional redirect>
"""
Field
Type
Default
Description
max_summary_tokens
u32
500
Max tokens in the bounded summary
narrative_schema
string
(default NOTES/PR/NEXT)
Template the sub-agent fills in inside <aide-summary> block
This section is load-bearing for aide-as-subagent mode. Without it, sub-agent output pollutes the frontier context. The runner wraps the task with instructions requiring the sub-agent to emit an <aide-summary> block conforming to the narrative_schema. The bounded summary is what aide wait returns to the calling agent, keeping the frontier context clean.
The read list is translated to .claude/settings.json permission grants and a WORKSPACE section in the task prompt. This solves sandbox restrictions in non-interactive claude -p mode, where the sub-agent would otherwise be unable to access files outside its own project directory.