Skip to content

slash-commands.md

Latest commit

 

History

History
188 lines (140 loc) · 8.67 KB

slash-commands.md

File metadata and controls

188 lines (140 loc) · 8.67 KB
Raw

Slash Commands

OpenAB registers Discord slash commands for session control and agent management. Most work in both guild threads and DMs — the exception is /auth, which is DM-only for security (see /auth below).

Commands

Command Description Requires active session?
/models Select the AI model via dropdown menu Yes
/agents Select the agent mode via dropdown menu Yes
/cancel Cancel the current in-flight operation Yes
/reset Reset the conversation session (clear history, start fresh) Yes
/auth Authenticate the backend agent via device flow (DM-only) No
/remind Set a one-shot delayed reminder to mention users/roles No
/export-thread Export thread/DM as .txt (default: last 100 messages) No

All responses are ephemeral — only the user who invoked the command sees the reply.

Platform Support

Platform Supported Notes
Discord (guild threads) ✅ Commands registered per-guild for instant availability
Discord (DMs) ✅ Commands registered globally; may take up to 1 hour to appear after first deploy
Slack ❌ Slack blocks third-party slash commands in threads; see slack.md

How They Work

/models and /agents

These read configOptions from the ACP initialize / session/new response and present them as a Discord Select Menu.

When the user picks an option, OpenAB sends session/set_config_option to the ACP backend.

Agent support varies:

Agent /models /agents
openab-agent ✅ Returns available models via configOptions in session/new response ❌
kiro-cli ✅ Returns available models via models fallback ✅ Returns modes (kiro_default, kiro_planner) via modes fallback
claude-code ❌ No configOptions emitted ❌
codex ❌ ❌
gemini ❌ ❌
cursor-agent ❌ (tracking: #493) ❌
copilot ❌ (tracking: #496) ❌

If the agent doesn't expose options, the user sees: ⚠️ No model options available. Start a conversation first by @mentioning the bot.

Backward compatibility: openab-agent returns configOptions in the session/new response (alongside sessionId). ACP clients that only read sessionId will continue to work — configOptions is additive. Clients that support /models should read configOptions[].options to populate the model picker. Each model option includes a provider field ("anthropic" or "openai") for routing.

Note: Discord Select Menus are limited to 25 items. If the agent returns more, only the first 25 are shown with a count of how many were truncated.

/cancel

Sends a session/cancel JSON-RPC notification to the ACP backend. This aborts in-flight LLM requests and tool calls immediately — no need to wait for the current response to finish.

/reset

Cancels any in-flight operation, then removes the session from the pool. The ACP process terminates once the last reference is released. The next message in the thread or DM will automatically create a fresh session.

This is equivalent to the sessions close + sessions new pattern used by OpenClaw ACPX.

What gets cleared:

  • Conversation history
  • ACP process and connection
  • Suspended session state (no resume after reset)

What is preserved:

  • Bot identity and system prompt (re-applied on next session creation)
  • Config settings in config.toml

/export-thread

Fetches the current Discord thread or DM history and returns a .txt file as an ephemeral follow-up. The transcript includes message timestamps, author names and IDs, message text, and attachment URLs.

Optional parameters (mutually exclusive — use at most one):

Parameter Type Description
limit Integer Export only the most recent N messages (1–5000)
since String Export messages after this message ID (right-click → Copy Message ID)
days Integer Export messages from the last N days (1–365). Rolling N×24h window from now.
all Boolean Export all messages (up to 5000)

If no parameter is provided, the last 100 messages are exported.

Examples:

/export-thread                              → last 100 messages (default)
/export-thread limit:500                    → most recent 500 messages
/export-thread since:1503744866100842698    → messages after this specific message
/export-thread days:3                       → messages from the last 3 days (rolling 72h)
/export-thread all:true                     → export all (cap 5000)

Constraints:

  • Only works in allowed Discord threads or enabled DMs.
  • Specifying more than one filter returns an error.
  • Very large exports may be truncated to fit Discord's attachment size limit.

Passing CLI Commands via @mention

In addition to slash commands, you can pass built-in CLI commands directly after an @mention:

@MyBot /compact
@MyBot /clear
@MyBot /model claude-sonnet-4

These are forwarded as-is to the ACP session as a prompt. Any command the underlying CLI supports in its interactive mode works here. This is the recommended workaround for agents that don't expose configOptions.

/remind

Set a one-shot delayed reminder that mentions users or roles in the channel after a specified delay.

Syntax:

/remind targets:<@user @role ...> message:<text> delay:<duration>

Parameters:

Parameter Required Description
targets Yes Space-separated @mentions (users and/or roles)
message Yes Reminder text
delay Yes Duration before firing: 1m to 30d (supports m, h, d and combinations like 1h30m)

Constraints:

  • Only humans can use /remind (bots are rejected)
  • Minimum delay: 1 minute
  • Maximum delay: 30 days
  • Maximum message length: 1800 characters
  • Maximum 5 active reminders per user
  • Maximum 10 mention targets per reminder (use a @role for larger groups)
  • @everyone and @here in messages are automatically neutralized (will not trigger mass mentions)
  • One-shot only (fires once, then removed)
  • Reminders persist across bot restarts (stored in $HOME/.openab/reminders.json)

Examples:

/remind targets:@Alice @Bob message:Review PR #42 delay:2h
/remind targets:@Reviewers message:Stand-up time delay:30m
/remind targets:@Charlie message:Check deployment delay:1d

When fired, the bot posts:

⏰ Reminder from @sender:
"Review PR #42"
cc @Alice @Bob

/auth

Trigger the backend agent's device-flow authentication. OAB executes the command defined in OPENAB_AGENT_AUTH_COMMAND, captures the device code URL from stdout/stderr, and relays it to the user as an ephemeral Discord message.

Flow:

  1. User runs /auth
  2. OAB executes $OPENAB_AGENT_AUTH_COMMAND (e.g. codex login --device-auth)
  3. OAB captures the device URL + code from the command's output
  4. OAB sends an ephemeral reply with the URL and code
  5. User opens the URL in their browser, enters the code
  6. The auth command exits successfully → OAB confirms "✅ Authentication successful!"

Requirements:

  • OPENAB_AGENT_AUTH_COMMAND environment variable must be set
  • The auth command must use OAuth device flow (print URL + code to stdout, then block until authorized)
  • No interactive stdin input required (headless-compatible)
  • Must be invoked in a DM with the bot (rejected in guild channels/threads for security)

Timeout: 14 minutes. If the user doesn't authorize within that window, the process is killed and the user is prompted to run /auth again. (Reduced from 15min to leave headroom for Discord's interaction token TTL.)

Behavior notes:

  • Only users in the allowed_users list can invoke /auth
  • Bot users are rejected — /auth is for human operators only
  • A 30-second URL-collection window waits for the auth command to print its URL. Slow-starting CLIs that take longer may show "no output".
  • Only one /auth flow can run at a time (single-flight). A second concurrent invocation is rejected with "already in progress".

Error cases:

  • OPENAB_AGENT_AUTH_COMMAND not set → immediate error message
  • Invoked by a bot user → rejected
  • Invoked outside a DM (in a guild channel/thread) → rejected for security
  • Auth command fails to start → error message
  • Auth command exits before printing a login URL (within the 30s window) → warning that no URL was produced, with a retry prompt
  • Auth command exits with non-zero → failure message with exit code
  • Timeout → process killed, retry prompt