openclaw/docs/reference
Peter Steinberger 139d7ab5e8
feat: react to prompts and replies with emoji in shared sessions (#161053)
People who can send to or suggest in a session can put emoji reactions on
any saved prompt or reply in the Control UI. Everyone reading the session
sees the chips update live, the agent sees each reaction as a System line
on its next turn (never woken for it), and reactions on prompts that came
from a channel are mirrored back to that channel as the bot's reaction.
Operators whose role only permits viewing see the chips without controls.

Storage is a same-version `session_reactions` companion table in the
per-agent database, installed on open like the session-sharing tables and
inert for older builds; reactions never touch transcript bytes. Two Gateway
methods (`session.reactions.set`/`list`) and one session event carry the
state, the hello frame projects the operator's session cap so the UI applies
the server's rule, and reaction reads run in the admitted history worker. Channel
mirroring keeps one bot reaction per emoji, dispatches in commit order, and
rechecks reactor authority and the source conversation before channel I/O.

The picker is a tapback-style pill below the message with six quick emoji,
pressed state for your own, keyboard navigation, and a custom entry that
applies a complete emoji the moment it lands (IME-safe, single emoji only).
Chips animate on change and list reactors with "You" first.

Live proof on a real Gateway with two identities and a QA channel; storage,
Gateway, UI, e2e and authority tests; published-2026.9.6 upgrade check.
2026-09-30 09:12:24 +00:00
..
database-schemas feat: react to prompts and replies with emoji in shared sessions (#161053) 2026-09-30 09:12:24 +00:00
full-release-validation feat(release): let a release lead record an exact-job flake instead of blocking publication (#161515) 2026-09-30 04:31:07 +00:00
session-management-compaction fix(compaction): prevent fallback from replaying previously compacted history (#160211) 2026-09-28 23:43:45 +05:30
templates improve(agents): stop agents refusing or arguing over requested work (#157447) 2026-09-24 16:16:21 -05:00
test fix(qa): Inspector collection fails when a pairing link is reused (#161540) 2026-09-30 01:48:56 -05:00
AGENTS.default.md improve(agents): stop agents refusing or arguing over requested work (#157447) 2026-09-24 16:16:21 -05:00
api-usage-costs.md
credits.md
database-schemas.md fix(state): observe foreign commits on the next database read (#156824) 2026-09-24 01:15:39 +00:00
device-models.md
full-release-validation.md fix(ci): preserve first failures through release validation 2026-09-23 18:27:19 -07:00
memory-config.md fix(bedrock): memory indexing fails under instance metadata throttling (#161164) 2026-09-29 20:39:49 +08:00
openclaw-ai.md docs: fix 12 concrete defects from the ux audit remainder (#144128) 2026-09-25 17:28:14 +08:00
prompt-caching.md fix(active-memory): recalls on claude-cli never reuse the prompt cache (#160768) 2026-09-29 04:52:53 +05:30
pull-request-review-flow.md fix(scripts): prevent host tooling drift from blocking PR workflows (#152078) 2026-09-18 14:34:18 -07:00
release-performance-sweep.md
RELEASING.md feat(release): let a release lead record an exact-job flake instead of blocking publication (#161515) 2026-09-30 04:31:07 +00:00
rich-output-protocol.md fix: resend inbound attachments referenced from chat history (#158720) 2026-09-26 21:02:02 +05:30
rpc.md
secret-placeholder-conventions.md docs: fix 12 concrete defects from the ux audit remainder (#144128) 2026-09-25 17:28:14 +08:00
secretref-credential-surface.md refactor: retire the TaskFlow Webhooks plugin (#158225) 2026-09-25 20:29:15 -07:00
secretref-user-supplied-credentials-matrix.json refactor: retire the TaskFlow Webhooks plugin (#158225) 2026-09-25 20:29:15 -07:00
session-management-compaction.md
test.md
token-use.md fix(codex): restore installed skills with catalog-backed models (#140422) 2026-09-24 14:59:15 -07:00
transcript-hygiene.md