Skip to content

Prepare Foreman 1.1.0-rc.1: working installs and persistent runtime - #5

Merged
fatihguner merged 1 commit into
mainfrom
codex/foreman-reliability
Sep 5, 2026
Merged

fatihguner merged 1 commit into
mainfrom
codex/foreman-reliability

Conversation

@fatihguner

Copy link
Copy Markdown
Owner

Summary

  • Make Foreman installable from a complete npm package and self-contained Claude/Codex bundles. The compiled OpenClaw adapter now registers 10 tools against the actual SDK.
  • Replace template-only persistence with shared, atomic profile, task and playbook state; repair typed content references and portable maintenance commands.
  • Prepare public 1.1.0-rc.1 artifacts, documented installation, package smoke checks and a manual npm trusted-publishing workflow.

Type of Change

  • Bug fix
  • Documentation
  • Infrastructure (scripts, CI, templates, config)

Layer Affected

Canonical content references, command registration, modes, implementation tracking, plugin adapters and bundles, npm CLI, maintenance scripts and release tooling. Most added lines are generated, committed plugin resources needed for installation directly from Git.

Testing

  • 23 automated tests pass locally; the suite also passed on Linux before release preparation.
  • Source validation checks 287 core Markdown files, 51 YAML documents, 4,152 typed references and 959 links with no errors.
  • Build, actual OpenClaw SDK typecheck, synchronized release metadata and workflow YAML validation pass.
  • The release tarball installs into a clean project, loads Claude/Codex resources, preserves existing files on repeated initialization and persists profile changes.
  • Real Claude command discovery and profile persistence were exercised; OpenClaw loaded all 10 tools.
  • GitHub CI checks Node 22/24 on macOS/Linux, generated bundle freshness and actual tarball installation before merge.

Release status

This is an explicit prerelease. Final live advisory quality acceptance remains pending: earlier exploratory answers included unsupported causal claims, invented targets and language drift. Stronger instructions are included, but the complete final regression has not passed. A live Codex model session is also unverified.

The release candidate uses npm's next channel when published. The stable latest channel remains 1.0.0. npm publishing requires a maintainer session or the documented trusted-publisher configuration; the public GitHub tarball is independently installable.

@fatihguner
fatihguner merged commit b6c6830 into main Sep 5, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant