@chat-adapter/slack

Patch Changes

• 7881071: Stop treating a bot reference as a mention when the rendered message shows it only as text. Slack sends an app_mention event for the bot's <@U…> token anywhere in a message's text field, including inside inline code or a code block, even when the message shows no mention. The adapter now reports isMention: false for these messages, so they no longer reach onNewMention.

  When an app_mention doesn't reference the bot's known user ID, the adapter treats it as a mention only if the rendered message shows a mention of some user outside code. This covers the W… IDs that Enterprise Grid workspaces can use for the bot. The same rule now applies when the bot user ID hasn't been resolved yet, where the adapter previously treated every app_mention as a mention.

  For messages whose blocks are Block Kit rather than rich_text, such as a header or section posted by an integration, the adapter now reads a bot token in text outside code as a mention when the blocks don't show the bot. This applies to both app_mention and message events, so the two events for the same message are classified the same way.

• 48bb905: Read the body of a Block Kit message from its blocks. When a message carries blocks, Slack renders them and uses text only as the notification fallback, so message.text and message.formatted could miss content an app or workflow placed in the blocks, including the request in a mention that reached onNewMention. When every block is a section, header, context, image, divider, actions, or table block, the text in those blocks now forms the body in block order, with tables in place and links from mrkdwn text in message.links. Messages with any other block type, including the rich_text blocks a Slack client sends, keep using text, so content the adapter can't render still reaches the message.

Minor Changes

• 7b7d73c: Add a visibleTo option to Button and LinkButton that lists the user IDs allowed to see the button. The Slack adapter renders it as visible_to_user_ids. Other adapters ignore it and show the button to everyone. It only hides the button, so action handlers still need to check who clicked.

• c70379b: Return every message ID from SlackAdapter.stream(). When a long reply rotates into several messages, the result now includes messageIds, which lists the messages of the reply in order. Its last entry is id. The new SlackStreamResult type describes the result.

• f923b1c: Add a blocks option to uploadSlackFiles so the message that shares the uploaded files can carry Block Kit blocks. Slack renders the files below the blocks. Passing both blocks and initialComment throws a TypeError, since Slack would silently ignore blocks.

• 9a83b22: Isolate Slack message concurrency and turn cancellation by installation to prevent cross-workspace bot token reuse. Adapters can now namespace dedupe, lock, queue, and turn keys with the optional Adapter.getConcurrencyScope() hook. In multi-workspace mode, messages queued under the old keys aren't migrated, and old and new workers don't exclude each other, so upgrade chat and @chat-adapter/slack together and restart all workers.

Patch Changes

• c70379b: Complete open task cards when a native Slack stream segment rotates, then replay the in-progress cards in the next segment. Retired segments now finish without Slack's error indicator. If Slack rejects the closing task cards, the adapter finalizes the segment without them and streams the rest of the reply as text only.

• b37a550: Preserve literal text and code formatting in incoming Slack messages and correctly parse email and phone links

• ec6edb8: Fix backward thread history reads returning older replies when a Slack thread spans multiple API pages. Backward reads now follow every conversations.replies page before selecting the newest messages, so long threads take one API call per page.

• c70379b: Cut Slack card text that goes past Block Kit's length limits, so a long value no longer makes Slack reject the message. This covers headers, section and muted text, subtitles, fields, link labels, button labels, image alt text, and select and radio options and placeholders. Cuts in mrkdwn text don't split a mention, link, or HTML entity, close a code fence left open, and keep the closing * of bold text.

Patch Changes

• 44423bd: Convert Slack special mentions (<!here>, <!channel>, <!everyone>) and user group mentions (<!subteam^ID|@handle>) to readable text in inbound messages.

Minor Changes

• ad90432: Add opt-in selection change events for Select and RadioSelect inputs in Slack modals

• 6adca36: Support Slack egress proxies across Socket Mode, response URLs, webhook forwarding, and guarded file downloads with adapter-scoped transport options.

Patch Changes

• 683eadc: Read the bot's mention from the message content instead of trusting the app_mention event type.

  Slack can deliver app_mention for a message whose bot-id reference only appears inside code, where it renders literally and is not an invocation. The adapter now classifies the content the way Slack renders it: a user element for the bot, or a <@U…> token in mrkdwn content, is a mention. Inline code (style.code), preformatted elements, rich-text text elements, link labels, and raw_text table cells render literally and are not mentions.

  When a message carries blocks, they model its body and the flattened text field is not consulted, because it loses the code/literal distinction. A message whose content was inspected and holds no mention reports isMention: false, even when the bot's display name appears as plain text. The synchronous parseMessage path classifies the same way, so the pre-edit snapshot of an edited message agrees with the edited message.

  An app_mention event whose content never shows the known bot id is still trusted, because Slack saw a mention under an id the adapter does not know (for example the W… id an Enterprise Grid workspace emits). When the adapter cannot identify the bot at all (botUserId unresolved), app_mention is trusted the same way and other messages stay undetermined.

  raw_text table cells now render literally, matching how they are classified: Slack reserves mentions and links for rich_text cells, so a <@U…> token in a raw_text cell no longer resolves to a display name.

• c21ccbc: Add an opt-in webhook option to propagate message, action, and slash command handler errors through waitUntil while preserving default behavior.

Minor Changes

• 51322dd: Decode the bot's own mention in incoming Slack messages. The adapter now resolves <@U_BOT> to the bot's display name (@<DisplayName>) the same way it resolves every other user mention, instead of leaving the raw user-ID markup in place, and sets isMention on the parsed message by detecting the bot's ID in the raw event text. This keeps message.text self-describing for downstream consumers (LLM prompts, classifiers) while preserving mention detection, which previously depended on the raw ID markup surviving in the text.

• d4a1f03: Rotate long-running native Slack streams before Slack expires them. Once a stream segment passes streamSegmentMaxAgeMs (default four minutes) the adapter finalizes it at the next paragraph break and continues the reply in a new message, closing and reopening code fences, repeating table headers, replaying open task cards and the plan title, and keeping the agent session in processing. A segment Slack already expired during an idle gap is recovered the same way instead of failing the reply.

• 2cc8cc3: Surface custom status text in the Agent messaging experience. startTyping and setAssistantStatus with a custom status now call the legacy assistant.threads.setStatus endpoint, whose compatibility bridge renders the text in the agent-session loading UX — instead of silently dropping the text and showing the generic "Working…" indicator. Clearing (empty status) still transitions the session to active via the Agent Sessions lifecycle.

Patch Changes

• 78021c0: pass workspace context to suggested prompt resolvers in Agent view

• 8b6d7f3: Reactivate Slack Agent Sessions when startTyping receives an empty status.

• c2b6bff: Use the Slack bot user ID for bot-authored messages when a bot profile is available.

• f485255: Harden webhook tenant isolation, require explicit Google Chat bot identity for reliable mention handling, use native Google Chat pagination, isolate Slack caches, and bound recording storage.

• 8fdaf4a: Stream post-and-edit fallback updates through Slack's markdown_text field instead of text, so live-updating messages render markdown while the stream is in progress (and get the 12,000-character ceiling rather than 4,000)

• 7609d8f: Validate external request targets before sending credentials, message content, or attachment requests.

Minor Changes

• 2ce2be0: Add Slack Agent Sessions lifecycle support, native stop cancellation through thread.signal, automatic session titles, and session stop/title-change events while preserving the legacy assistant_view compatibility path.

Patch Changes

• 864d922: Keep alert attachment content on normalized Slack messages. Attachments that aren't link unfurls now contribute their pretext, title (linked to title_link when present, with the URL also surfaced in message.links), text, and fields instead of being dropped; fallback fills in when nothing else on the attachment carries content. Matching how Slack renders these fields, they are treated as plain text unless listed in the attachment's mrkdwn_in array, so literal *, _, and backticks in alert text survive normalization. Tables inside attachment blocks now stay adjacent to their attachment's text.

  Because attachment content is part of message.text, mention detection and onMessage pattern handlers see it too: an attachment that quotes the bot's mention routes to onNewMention, and patterns match alert text. Handlers that should ignore other integrations' alerts can check message.author.isBot.

• 7c26965: prevent attachment downloads from sending credentials to untrusted hosts

• e71bfea: Preserve the first line of incoming Slack code blocks when extracting message text and formatted content.

  Only paired triple-backtick fences become code blocks: an unpaired fence, a fence inside inline code or a <…> token, and a fence on a quoted line all stay literal text, matching how Slack renders them. Bold and strikethrough rewriting no longer touches fenced code content, and text following a closing fence can no longer turn into a blockquote, heading, or list.

• b6fa24c: guard attachment downloads across the remaining adapters

  Slack, Discord, and WhatsApp attachment downloads now go through the shared guarded downloader: private and internal addresses are refused (as URL literals, through DNS resolution, and after redirects), responses are capped at 25 MB, and downloads time out after 30 seconds. Slack sends the bot token only on hops to trusted Slack origins, and WhatsApp keeps its access token on Meta's media hosts and the configured Graph origin. Telegram enforces the same size cap and timeout with the Web Fetch API so downloads keep working in runtimes like Cloudflare Workers.

  downloadAttachment in @chat-adapter/shared now resolves headers per hop (pass a function to control what each redirect target receives), forwards the resolved headers to custom transports, and accepts an onResponse hook to reject unexpected final responses before the body is read.

Patch Changes

• 764e475: preserve pasted Slack tables in parsed message content: table and data_table blocks become structured table nodes in message.formatted and tab-separated text in message.text. Table cells render mentions, channels, and links through the same mrkdwn converter as body text, unfurl and app attachments are excluded, date cells without a fallback are formatted from their timestamp, and tables pasted above the message text stay above it. Headerless pasted tables get an empty header row so GFM re-serialization doesn't promote the first data row to a header. SlackEvent.blocks is now typed with the exported SlackMessageBlock interface.

Minor Changes

• 4ac0455: Add message update and delete lifecycle callbacks, with Slack message_changed and message_deleted dispatch support.

Patch Changes

• 6f0d2f0: Resolve outgoing @name mentions on the Slack native streaming path so streamed responses mention users consistently with the post-and-edit fallback. Committed renderer text is resolved incrementally, keeping fenced code literal and preserving the existing ambiguity semantics.

• 4cc3445: Bound the length of bracketed URLs parsed from message text in the link-unfurl fallback, avoiding a quadratic scan on adversarial input. Valid links are unaffected.

• c311827: Preserve the Slack channel ID when converting labeled channel tokens (<#C123|general> now becomes #general (C123)) so agents can pass the ID to channel tools, and normalize the commonly hallucinated <label|url> link order before Markdown conversion

Minor Changes

• 0153a39: Add DateInput and NumberInput modal children. The Slack adapter renders them as a datepicker and a number_input, the Teams adapter as Input.Date and Input.Number, and both submitted values arrive in event.values as strings.

  Teams submit values that arrive as JSON numbers are now stringified into event.values instead of being dropped. This fixes Input.Number, but applies to any numeric value a Teams dialog submits — a key that was previously absent from event.values will now be present as a string.

Minor Changes

• bb7cd12: Expose sender email addresses on normalized incoming Slack message authors. message.author.email is populated from the same cached users.info lookup used for display names and requires the users:read.email scope; without it the field stays undefined.

• 907450d: Enterprise Grid fixes:

  • handleOAuthCallback now handles org-wide installs (is_enterprise_install): Slack returns team: null for these, and the installation is now keyed by the enterprise ID — the same key webhook token resolution looks up — instead of failing. The result includes enterpriseId and isEnterpriseInstall, and SlackInstallation records both.

  • Socket mode now resolves per-installation tokens for events, slash commands, and interactive payloads in multi-workspace deployments (matching the HTTP webhook path), and no longer drops enterprise_id / is_enterprise_install / is_ext_shared_channel from event payloads.

  • The user profile cache and display-name mention reverse index are now scoped by installation in multi-workspace deployments, so profiles fetched with one workspace's token no longer bleed into another and mentions can no longer resolve to a same-named user from a different workspace. Existing cache entries repopulate on first lookup (single-workspace keys are unchanged). withBotToken accepts an optional { installationId } so proactive/cron posts outside webhook handling scope these caches too.

  • API calls made while handling an event from an org-wide install now pass the event's team_id explicitly, as Slack requires for workspace-scoped methods (conversations.list, usergroups.*, …) on org tokens. When an event carries a context_team_id (shared channels hosted on an "away" workspace), channel-addressed calls echo it back as client_context_team_id.

  • Retried event deliveries (x-slack-retry-num, socket retry_num) are dropped when the original delivery was already dispatched, using an event_id marker in the state adapter (24-hour TTL). Events whose first delivery never arrived are still recovered via the retry.

  • Bare @W… mentions in outgoing messages are now recognized as raw Enterprise Grid user IDs (previously only @U… was), so they render as real mentions instead of being treated as display names.

  • Event token resolution now prefers the envelope's authorizations[0] — Slack's documented location for the event's installation identity — over the top-level team_id/enterprise_id, which can name a different workspace for Slack Connect shared-channel events. Top-level fields remain as a fallback.

Patch Changes

• 80def3a: Add optional isSystem field to the normalized message Author type to distinguish platform-generated messages from humans and bots. The Slack adapter now sets isSystem: true for messages authored by Slack's reserved USLACK user (e.g. "@user archived the channel" notifications in DMs), so consumers no longer need to hard-code Slack-specific user IDs.

• 92530dd: Return a replyable Slack thread ID from channel.post() by using the posted top-level message's timestamp as the thread root.