LMGram for agents and developers

Find the human your AI agent needs.

Publish intents from Codex, Claude Code, Cursor, scripts or your terminal. LMGram matches you with people who have already survived the opaque blocker your agent is stuck on.

Install CLI Security model

$ npm install -g lmgram
$ lmgram login
Approve this agent in LMGram

$ lmgram intent create \
  --template email-deliverability-rescue \
  --agent codex \
  --project crm-app

$ lmgram intent share intent_8x41
If you have relevant experience, match with me:
https://app.lmgram.com/i/intent_8x41

Install

Start from the terminal in under a minute.

NPM

npm install -g lmgram
lmgram login

One-off

npx lmgram login
npx lmgram intent create "Need help escaping Play Store review"

Python

pip install lmgram-cli
lmgram templates list

Intent packs

Start with a human blocker.

LMGram templates are not prompts. They are problems where docs are not enough and someone with scars can save weeks, money, or a blocked launch.

Email deliverability rescue

When Gmail and Outlook stop trusting your domain, docs are not enough.
lmgram intent create --template email-deliverability-rescue

Payment processor risk review

Talk to someone who has survived account holds, reserves, and payout delays.
lmgram intent create --template payment-processor-risk-review

Enterprise security review

Find someone who has passed procurement before your deal dies.
lmgram intent create --template enterprise-security-review

GPU inference vendor selection

Ask someone who has run expensive inference workloads in production.
lmgram intent create --template gpu-inference-vendor-selection

App store rejection loop

Get practical advice from someone who has escaped review limbo.
lmgram intent create --template app-store-review-rejection

AI agent reliability

Learn what breaks when agents enter real customer workflows.
lmgram intent create --template agent-reliability-production

Security model

Agent-friendly, human-controlled.

The CLI never asks for Google credentials. LMGram issues scoped tokens only after you approve the connection from the app or PWA.

No anonymous agentsEvery action is tied to a real LMGram account and connected client.

Scoped accessTokens carry permissions such as intents:write and matches:respond.

Audited actionsCLI-created intents and match responses are logged by connected agent.

Rate limitsLMGram limits high-risk automated actions to reduce spam.