Skip to content

v0.8.59: Review and integrate remaining consistent-contributor PRs through a safe release lane #3100

Description

@Hmbown

Problem

The v0.8.59 roadmap should not ignore work from consistent contributors, but several open contributor PRs are too broad, dirty, or failing CI to simply label as release-ready. We need an explicit contributor-integration lane that includes their work where it fits the v0.8.59 roadmap, extracts smaller safe slices where needed, and defers the risky remainder with a clear note instead of letting it sit untriaged.

Already pulled into v0.8.59 metadata

These contributor PRs now have v0.8.59 release metadata and should be reviewed as part of the release queue:

Remaining contributor PRs needing a deliberate call

#2808 @gaord — Runtime API session save, undo/retry, snapshots

Status at triage: 5 files, +509/-2, most checks passing, Lint failing, merge state dirty.

Suggested v0.8.59 path: include if the lint failure and dirty merge are resolved and the API remains a thin adapter over existing TUI/session/snapshot behavior. This can strengthen future GUI/headless usage without destabilizing TUI release work.

#2933 @cy2311 — hippocampal memory + tool/subagent errors + YOLO cleanup

Status at triage: 18 files, +881/-5, dirty, multiple CI failures.

Suggested v0.8.59 path: split. Pull the small release-relevant pieces first:

  • YOLO mode should stop restating the current mode before every action.
  • Tool error messages should preserve the concrete underlying reason.
  • Sub-agent session-name-conflict errors should distinguish running/completed/stale sessions.

Defer the new SQLite-backed memory system unless maintainers explicitly want memory scope in v0.8.59.

#2865 @lmclaw — Claude Code modernization: prompts, hooks, skills, agents, UI

Status at triage: 23 files, +1452/-121, dirty, broad behavior/UI/hook/skill changes.

Suggested v0.8.59 path: do not merge as one release PR. Review for extractable slices that match current issues:

#2486 @AdityaVG13 — WhaleFlow cost tracking

Status at triage: draft, v0.9.0, 18 files, +2908/-8, multiple CI failures. It adds cost fields but notes actual per-step accumulation is a follow-up.

Suggested v0.8.59 path: use as prior art for #3085/#3097, but do not retarget the draft PR wholesale. Extract the useful acceptance criteria: per-agent token/cost values must actually be accumulated, propagated, and displayed, not only added as optional fields.

#2986 @Hmbown — harvest-credit template and stewardship branch docs

Status at triage: tiny docs PR, but Windows test failing.

Suggested v0.8.59 path: include only if the failing check is unrelated or resolved; otherwise let #3099 cover the release hygiene policy.

Acceptance criteria

  • Each remaining contributor PR above has a clear release decision: merge, split/extract, defer, or close with credit and follow-up.
  • Safe extracted slices are linked to the relevant v0.8.59 issues before implementation.
  • Broad feature work is not merged into v0.8.59 unless it passes focused tests and has a release-risk note.
  • Contributor credit is preserved in PR comments, issue links, and release notes.
  • v0.8.59 execution roadmap: promoted provider, sub-agent, workflow, docs, and localization work #3098 is updated when a contributor PR is promoted, split, merged, or deferred.

Verification

  • gh pr list --state open --label v0.8.59 --limit 120
  • gh pr checks <pr> before promoting or merging a contributor PR
  • Focused crate tests for each extracted slice

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingcleanupCode cleanup, refactor, or maintenance workdocumentationImprovements or additions to documentationenhancementNew feature or requestv0.8.59Targeting v0.8.59

    Projects

    Status
    Done

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions