This repository was archived by the owner on Jun 16, 2026. It is now read-only.
v0.2.0
What's Changed
- fix(security): backport security fixes from openclaw main by @arthyn in #63
- fix: sanitize user messages to prevent role tag prompt injection by @patosullivan in #64
- fix: pin api-beta to S3 endpoint fix by @arthyn in #65
- feat: use skill's cookie cache for faster auth by @arthyn in #66
- add restart internal ring ships to ci by @yapishu in #68
- feat: use bot meta and rate limit by @arthyn in #67
- CI: fix GCP auth and OVH context by @yapishu in #70
- feat: improve approval UX and fix DM allowlist security holes by @patosullivan in #69
- feat: pass botProfile to all monitor send calls by @arthyn in #71
- security: skip failing tests by @arthyn in #72
- fix: use senderShip for DM OriginatingTo routing by @arthyn in #73
- feat: add /tlon-version slash command by @arthyn in #75
- chore: track version.generated.ts for copied source installs by @arthyn in #76
- fix: handle bot-meta author format in channel and DM handlers by @arthyn in #74
- Cleanup approval flow + emoji reaction approvals by @patosullivan in #77
- monitor: dont reset on self message (just continues loop) by @arthyn in #78
- ci: split into PR tests and master deploy workflows by @arthyn in #79
- fix: remove duplicate ship names in approval notification messages by @patosullivan in #80
- fix: wire ownerShip into core auth via resolveAllowFrom/formatAllowFrom by @patosullivan in #81
- Bump version to 0.2.0 by @arthyn in #82
Full Changelog: https://github.com/tloncorp/openclaw-tlon/commits/v0.2.0