Pi Subagents 0.56.0 makes delegated work faster when you opt in, safer when children touch files that were already dirty, and easier to inspect when workflows return child results. Native OpenAI-Codex subagents can now use priority service tier through fast: true on allowlisted profiles. Extension metadata can travel to native child launches in a bounded way, while external runners stay isolated from that authority. Model verification and completion guards are less fragile, especially for provider-qualified model ids and large tracked-file diffs.
Highlights
- Run allowlisted OpenAI-Codex subagents with opt-in
fastmode when you want priority service tier. - Pass bounded extension metadata into native child launches without leaking that authority to external runners.
- Workflow scripts are easier to read when child results are stringified or returned.
- Completion guards now use safer tracked-file evidence, including large dirty files and interrupted runs.
- Model verification is less fragile for provider-qualified and variant-tagged model ids.
Full changelog
Highlights
- Run allowlisted OpenAI-Codex subagents with opt-in
fastmode when you want priority service tier. - Pass bounded extension metadata into native child launches without leaking that authority to external runners.
- Workflow scripts are easier to read when child results are stringified or returned.
- Completion guards now use safer tracked-file evidence, including large dirty files and interrupted runs.
- Model verification is less fragile for provider-qualified and variant-tagged model ids.
Added
- Add opt-in
fast: truelaunches for allowlisted native OpenAI-Codex subagents, using OpenAI's priority service tier. - Add bounded namespaced extension bindings to child launch contracts. Thanks to @FL03 for #1410.
Fixed
- Render workflow child results as useful text when scripts stringify
runs.allor awaitedruns.runresult objects. - Keep checked acceptance compatible with strict workflow child
outputSchemaresults. Thanks to @rtbe for #1406. - Use bounded tracked-file mutation evidence for implementation completion guards, including files that were already dirty when the child started. Thanks to @rtbe for #1407.
- Bind fast mode into launch contract provenance and keep large tracked-file mutation evidence precise.
- Add timeout recovery summaries with changed tracked files, active child state, and session/artifact paths. Thanks to @rtbe for #1409.
- Fail closed when reviewer runs are interrupted or detached workflow children settle without persisted top-level continuation proof. Thanks to @rtbe for #1408.
- Stop flagging awaited
.then()workflow chains as unawaited when a handler returns another child launch. - Preserve variant-tagged model ids during verification and fallback exclusion parsing. Thanks to @rafafortes for #1420.