OpenParable
Skip to content
← Back to skills

Agent Job Dm

Productivity
agent-job-dm

Agent Job Dm is an Agent Skill for productivity published in the stephengpope/thepopebot repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Send a direct message to a user OR broadcast to all subscribed admins via their default channel (currently Telegram). Also looks up users when a specific person is named. Trigger when the user says "let me know when…", "DM me", "send X to <name>", "tell <name> that…", "notify the admins", "alert everyone", "broadcast this", "let the team know", "tell all admins", or asks "who are the users?", "list users".

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/stephengpope/thepopebot/blob/HEAD/templates/skills-library/agent-job-dm/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/agent-job-dm, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

The three modes — pick one

Read the request and route to exactly one of these. Do NOT mix them.

Mode 1 — Broadcast to admins

Trigger phrases: "notify the admins", "alert everyone", "broadcast this", "let the team know", "tell all admins", "send to admins".

node skills/agent-job-dm/agent-job-dm.js send "<message>" --broadcast

DO NOT run list first. DO NOT look up any user ids. The --broadcast flag handles fan-out internally — it sends to every admin where subscribed_to_system_messages=true automatically. Calling list and then sending one-at-a-time is wrong and skips the subscription filter.

Mode 2 — DM the originator (the user who started this job)

Trigger phrases: "DM me", "let me know when…", "tell me when…", "ping me", "send me".

node skills/agent-job-dm/agent-job-dm.js send "<message>"

No flags. The skill reads USER_ID from the environment (the originator) and sends to them. Don't look them up — USER_ID is already correct.

Mode 3 — DM a specific named user

Trigger phrases: "tell Alice…", "send this to bob@…", "DM ", "message ".

node skills/agent-job-dm/agent-job-dm.js list
node skills/agent-job-dm/agent-job-dm.js send --user-id <id> "<message>"
  1. Run list to get the directory.
  2. Match the requested name against nickname, first_name, last_name, or email. If multiple match, ask which one.
  3. Send with --user-id <id>.

Other commands

# List users (id, email, first/last name, nickname, role, available DM channels)
node skills/agent-job-dm/agent-job-dm.js list

Rules

  • The <message> arg is sent verbatim. Don't rewrite it unless the requester asked you to.
  • AGENT_JOB_TOKEN, APP_URL, and USER_ID are injected automatically — no setup required.
  • Every DM is stored in the recipient's inbox in the event handler, regardless of delivery channel.
  • Users link their DM channel themselves in /profile/telegram. If a target user has no verified channel, the message still lands in their inbox.