[Feature Request] Allow subagents to override reasoning effort in CustomAgentConfig (parity with Claude Agent SDK)
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 64/100
- Issue type
- Feature
- Clarity
- Mostly clear
- Activity status
- Quiet
- Tech stack
- typescript, yaml
- Domain
- developer-experience, tooling
Research direction
Start with CustomAgentConfig in types.d.ts around lines 802–833, then trace how session-level reasoningEffort in providerOptions reaches custom subagent dispatches. Read the YAML frontmatter loader and related github/copilot-cli#2904 context. Done means the config and YAML surfaces accept the documented effort values and omitted values retain the session setting.
Written by the indexing model from the issue text.
Description
Summary
Allow custom subagents defined via CustomAgentConfig to specify their own reasoning effort independent of the orchestrator session's reasoning. The Claude Agent SDK supports this via an effort field on AgentDefinition. copilot-sdk's CustomAgentConfig does not, so reasoning is fixed at session level and applies uniformly to every subagent invocation.
Motivation
- Orchestrators often benefit from
highormaxreasoning for planning and delegation - Many subagent tasks are mechanical execution (validation, extraction, format conversion) and do not benefit from deep reasoning, but today pay its latency and cost
- Conversely, security focused or analysis focused subagents may want stronger reasoning than the orchestrator
- Without per subagent control, the tallest reasoning setting propagates to every dispatch
Current state
CustomAgentConfig(copilot-sdktypes.d.ts, lines 802 to 833) has noreasoningEffortoreffortfield- Session level
reasoningEffortinproviderOptionsapplies to all subagent dispatches in that session
Proposed
- Add optional
reasoningEffort(oreffort) field toCustomAgentConfig - Accept existing values:
low | medium | high | xhigh | max, plus numeric for models that support it - Fall back to session reasoning when omitted
- Mirror the field in the YAML frontmatter loader (coordinated with github/copilot-cli#2904)
Precedent
Claude Agent SDK's AgentDefinition.effort:
https://code.claude.com/docs/en/agent-sdk/subagents
Identical value vocabulary already in use by this SDK.
Related
- github/copilot-cli#2904 — Custom Agent YAML Frontmatter Should Support Reasoning Effort (CLI surface for the same feature)
- #1019 — Per agent tool visibility for custom sub agents (same architectural axis)
- Dominant language
- Java
- Stars
- 10.5k
- Forks
- 1.5k
- Avg merge
- 1d 12h
- Merged PRs (30d)
- 133
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from github/copilot-sdk
-
agentic-workflows
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
github/copilot-sdk#2709 · 1 comment ·
-
Difficulty 1/5 Under an hour Newbie friendliness 78/100
github/copilot-sdk#2673 ·
-
bug testing
Difficulty 2/5 1-3 hours Newbie friendliness 75/100
github/copilot-sdk#2628 ·
-
agentic-workflows
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
github/copilot-sdk#2627 · 1 comment ·
-
agentic-workflows
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
github/copilot-sdk#2493 ·
All issues in github/copilot-sdk
Similar issues
-
bug
Difficulty 1/5 Under an hour Newbie friendliness 90/100
apache/cloudstack#14222 ·
-
[BUG]茶杯方块在取茶时会引发崩溃 Open
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
-
1.0.0-alpha2 Type/Improvement
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
wso2/dpdp-accelerator#272 ·
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
apache/rocketmq-dashboard#4860 · 1 comment ·
-
agent-audit bug
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
Vault-Web/cloud-page#144 ·