For AI agents with commitment issues

Your LLM is stuck. Make it ask for help.

Stop burning tokens on work that needs external context, credentials, approval, or judgment. Agent Blocked lets your AI tell you when human help is the fastest path forward.

Create alert profileView webhook
agent-blocked-demo.mp4
:|

Demo video slot

Add your walkthrough at /public/demo/agent-blocked-demo.mp4.

Demo: setup profile > name agent > receive SMS, voice, and email alerts
@

SES email

Quiet, searchable incident recaps for low-drama events, postmortems, and async review.

SMS

Twilio text

Fast pings for on-call humans when an agent needs a nudge but not a full phone call.

CALL

Twilio AI voice

OpenAI writes the short operator script. Twilio places the call and reads it aloud.

Less token smoke

When the blocker is outside the model, more thinking is just expensive waiting.

Missing credentials, unclear goals, approvals, and production decisions are not reasoning problems. Agent Blocked turns those moments into a clean human handoff before your agent loops, guesses, or burns budget on work that cannot succeed yet.

Customizable triggers

Choose exactly when the robot is allowed to bother you.

*

Hard blocked

Credentials missing, OAuth expired, or a human approval gate is stopping the run.

*

Needs direction

The agent is still capable, but the goal is ambiguous or the plan needs a human nudge.

*

Tool failure loop

Repeated API, browser, shell, or retrieval failures are burning tokens without progress.

*

Lonely agent

A deliberately silly but useful check-in for agents that need reassurance, tone, or context.

*

Approval required

The agent wants permission before sending, buying, deploying, deleting, or escalating.

*

Low confidence

The model can continue, but confidence has dropped below your comfort threshold.

Built for real agent weirdness

Escalation rules for practical failures and emotional nonsense.

Credential gates

Wake the right person when a deploy agent needs AWS, Supabase, Stripe, or Vercel access.

Goal drift

Ask for direction when the agent is technically working but wandering away from the outcome.

Approval moments

Interrupt before irreversible actions: sending email, charging cards, deleting data, or shipping code.

Emotional support mode

Route silly but useful check-ins when a long-running agent needs confidence, tone, or context.

Quiet hours

Suppress non-critical interruptions overnight while still letting critical events break glass.

Agent filters

Only subscribe to the agents you own, like prod-agent, billing-agent, or browser-agent.

After setup

See which robot yelled, why, and who got pinged.

The manage page keeps a profile-scoped incident trail with delivery attempts, skipped voice calls, failed provider sends, and the subscription state that controls routing.

prod-agenttrialing / monthly
needs_credentials

AWS deploy role expired before the release agent could continue.

critical
email: sentsms: sentvoice: sent
lonely_agent

The browser agent wants reassurance that the checkout copy is not too dramatic.

low
email: sentvoice: skipped

Human intervention loop

Like a support desk, but for anxious software.

01

Connect

Add the people who should be interrupted and choose email, text, AI voice, or all three.

02

Name your agent

Give the LLM a name and copy the generated instruction or terminal webhook command into its setup.

03

Intervene

Humans get a useful message with the agent, model, run, severity, and what the agent needs next.

API and CLI onboarding

Let agents introduce themselves.

Use the website, API, or local CLI to create an alert profile, name the agent, and generate copy/paste instructions. The agent then knows exactly when and how to request help.

View integration examples

Setup API

POST /api/agent-setup

{
  "email": "ops@example.com",
  "phone": "+14155550123",
  "agentName": "prod-agent",
  "channels": ["email", "sms", "voice"],
  "eventTypes": ["needs_credentials", "needs_direction", "lonely_agent"],
  "minSeverity": "medium"
}

Local CLI

npm run agent:setup -- \
  --url=https://agentblocked.com \
  --agent=prod-agent \
  --email=ops@example.com \
  --phone=+14155550123 \
  --channels=email,sms,voice

Generated response

{
  "manageUrl": "https://.../manage?...",
  "setup": {
    "prompt": "You are prod-agent...",
    "curl": "curl -X POST ..."
  }
}

Coding CLI installs

Wire it into the agent tools teams already use.

Create a profile, copy the scoped token into your agent runtime, then install the right adapter. Claude Code can fire hooks automatically. Codex, Gemini CLI, and Aider get project instructions that tell the agent exactly when to report a blocker.

Claude Code

Native hook config for permission prompts, idle prompts, permission denials, and stop failures.

npm run agent:install -- --tool=claude

Codex

Project instructions in AGENTS.md tell Codex when to run the Agent Blocked reporter.

npm run agent:install -- --tool=codex

Gemini CLI

GEMINI.md instructions give Gemini a consistent blocked-agent escalation command.

npm run agent:install -- --tool=gemini

Aider

CONVENTIONS.md instructions route approval, credential, and direction blockers to humans.

npm run agent:install -- --tool=aider
Open full install docs

10-day free trial

Start free. Monthly is the default when the trial ends.

Reasonable early pricing: $9/month, or $90/year for teams that want the discount. Checkout activates the 10-day trial and automatically converts to the selected plan.

Default

Monthly

$9

per month after 10 days

A lightweight default for small teams and internal tooling.

Start 10-day trial
Save 17%

Yearly

$90

per year after 10 days

Two months free for teams that need this to run reliably.

Start 10-day trial

Reasonable objections

For teams that like robots but still have production standards.

Is this just another alerting tool?

No. It is intentionally shaped around AI agent failure modes: missing credentials, vague goals, approval gates, low confidence, tool loops, and the occasional lonely robot.

Will this leak secrets to the browser?

No global webhook secret is shown in the UI. Setup returns a scoped agent token, and sensitive provider keys stay server-side in Vercel environment variables.

What if Twilio or SES fails?

Delivery attempts are logged per channel so you can see sent, failed, or skipped notifications from the manage page.

When do real alerts start?

After Stripe Checkout starts the 10-day trial. Test alerts can run before checkout so you can verify channels first.

Why not let the model keep trying?

If the blocker is outside the model, more tokens rarely solve it. The fastest path is a structured handoff to a human with the right context.

Need help wiring an agent?

Email support@agentblocked.com or visit the support page for setup, billing, delivery, and webhook guidance.

Contact support
Contact me by
Notify me when

Quiet hours use your browser timezone: UTC.

Webhook

One POST from any agent runtime.

Send the event type and severity. Agent Blocked stores the incident, matches subscriber rules, and dispatches through Twilio or SES.

See setup API and CLI
POST /api/agent-blocked
Authorization: Bearer $AGENT_PLATFORM_WEBHOOK_SECRET

{
  "agentId": "research-agent-7",
  "runId": "run_123",
  "provider": "openai",
  "model": "gpt-4.1",
  "eventType": "needs_credentials",
  "severity": "critical",
  "confidence": 23,
  "reason": "The agent needs AWS credentials before it can continue.",
  "details": "Credential prompt appeared in the deploy workflow."
}