Skip to main content
Use this when you are about to send a DM, email, sales note, founder update, investor message, partnership ask, or follow-up. The important part is not whether the message sounds good in isolation. The important part is how the recipient may privately interpret it.

​
What to give the agent

I want to send a DM. Use Yomira before I send it.

Recipient:
- Who they are:
- What they liked, replied to, or asked for:
- What they probably care about:
- What they may be skeptical of:

My context:
- Who I am:
- What I am building:
- What relationship I already have with them:
- What I want them to do:

Candidate message:
"""
Paste the exact message here.
"""

​
Agent workflow

  1. Extract the exact message as the artifact.
  2. Build a small but specific audience: the recipient, adjacent buyers, skeptical peers, and people with similar context.
  3. Call Yomira with the message and context.
  4. Return reaction clusters, raw voices, and concrete edits.
  5. If there are multiple message variants, simulate every variant.

​
API shape

curl -s -X POST "https://tryyomira.com/api/simulate" \
  -H "content-type: application/json" \
  -H "x-api-key: $YOMIRA_API_KEY" \
  -d '{
    "objective": "Decide whether to send this DM and how to revise it.",
    "artifact": {
      "type": "dm",
      "content": "Paste the exact DM here."
    },
    "audience": {
      "description": "The specific recipient plus similar potential users who saw the post, understand AI agents, and may care about reaction simulation."
    },
    "simulation": {
      "mode": "fast",
      "target_n": 60,
      "max_agent_voices": 12,
      "max_output_tokens": 30000
    }
  }'

​
Good output

Ask the agent to return:
  • send / revise / do not send
  • the biggest trust gap
  • the strongest positive interpretation
  • the most likely reason they ignore it
  • three representative raw voices
  • a revised message if revision is needed