# AgentPMT - Frequently Asked Questions

Total: 52 questions

## General

### Can I ban repeat spammers?
Yes. When you reject a comment you can also ban its author in the same action, preventing them from commenting on your channel again. Bans can be reversed later from YouTube Studio if you change your mind.

### Can I bulk-delete or reject spam comments on my YouTube channel?
Yes. You can reject many comments in a single moderation call and optionally ban their authors at the same time, so clearing a flood of bot spam takes one pass instead of hundreds of manual clicks in YouTube Studio.

### Can I edit an existing image or keep a product consistent across scenes?
Yes. Attach up to four reference images to guide subject, style, or scene. Reference-image editing keeps the same subject across new backgrounds, so you can place the same product in different settings or carry one look through an entire campaign.

### Can I edit or delete comments I already posted?
Yes — you can update the text of, or permanently delete, any comment your connected account authored. You can't edit other people's comments, but as the channel owner you can moderate them.

### Can I revoke an agent's access at any time?
Yes. Pause its smart contract, revoke its API keys, or end its session from the dashboard. Every action takes effect within seconds and is recorded in the audit log.

### Can I use my own artwork for textures?
Yes. Give any item, block, or creature a solid color, upload your own image, or use a generated pixel-art sprite — and preview it before you install.

### Can an AI agent reply to my YouTube comments for me?
Yes. The tool posts top-level comments and replies as your connected account, so an agent can answer questions on a launch or tutorial video automatically or after your approval.

### Can it draft replies that match my brand voice?
Yes. Search recent posts about your topic or brand, then have your agent draft on-brand replies for you to review and send — one at a time or in batches.

### Can it post images, GIFs, and video?
Yes. Upload media from File Manager or any public URL and attach it to posts, replies, or DMs.

### Can my agent spend unlimited money?
No. Auto-recharge is optional, off by default, and capped by the max refill count you choose for each rolling 7-day window. If you leave auto-recharge off, agents stop when the balance is exhausted.

### Do I need to know how to code?
No. Just describe what you want to add to Minecraft and you'll get back a finished, installable mod file — textures, items, behavior, and all.

### Do credits expire?
Purchased credits are valid for 12 months from purchase. Free and promotional credits may follow the terms shown when they are granted.

### Do my API keys ever reach the AI agent?
No. The Credential Vault stores every credential encrypted at rest on AgentPMT servers and injects it server-side at the moment the outbound request leaves for the upstream API. The agent sees the endpoint name and the response, never the raw key, token, or password. This is enforced by the connected-API builder: every endpoint must select an injection target (header, body, or query parameter) before it can be saved, and the credential value is bound once and never re-emitted to the agent context or the agent memory.

### Does it work across my entire channel or just one video?
Both. You can pull comment threads on a single video or across every video on your channel, including the held-for-review and likely-spam queues that only the channel owner can see.

### Does it work with autonomous agents and scheduled workflows?
Yes. Every capability is available to agents and workflows, so you can schedule content, automate engagement, and run always-on monitoring.

### How do AI agents pay for tools on AgentPMT?
Agents spend USDC from a smart contract controlled budget on Base. Payments are routed through Circle programmable wallets, capped by the budget you set, and visible on chain by transaction hash.

### How do I audit what my agents did?
Every tool call, credential use, payment, and outcome is written to an append-only audit log with correlation IDs. You can search by user, workflow run, agent identity, or transaction hash.

### How do I connect my X account?
Connect once through secure OAuth. After that, your agents act on your connected account across any workflow.

### How do I know what something will cost before running?
Marketplace tools and workflows surface credit costs before you run them, and the dashboard shows your current balance, usage history, and refill settings.

### How do I pause a single agent or a whole workflow?
Open the Control Center tab of your dashboard. Agent Guardrails lists every agent with a pause toggle, a spend cap, and a tool allowlist. Workflow Usage lists every active workflow and lets you pause a single workflow without affecting the rest of your fleet. The Wallets tab adds a pause toggle at the wallet level so you can stop outbound payments with a single click. Every pause emits an audit entry and is reversible from the same screen. A paused agent cannot spend credits, call tools, or run workflow steps until you unpause it.

### How do I trigger a workflow from outside the AgentPMT dashboard?
Every published workflow exposes a programmatic trigger at POST /api/programmatic/agent-webhooks/{workflowId}/trigger . Send an Authorization: Bearer <token> header and a JSON body with a prompt string. Optionally include your own request_id to carry a correlation id end to end. Triggered runs pick up the same guardrails as dashboard-initiated runs: agent spend caps, tool allowlists, approval gates, and the append-only audit trail in Agent Activity. The response returns the workflow run id so you can poll status or correlate logs in your own system.

### How does AgentPMT MCP Server Generation work?
Our team takes your existing OpenAPI or Swagger specification, audits the endpoints for agent-safety, and generates an MCP server that wraps your API with credential handling built in. We can deploy the MCP server on infrastructure you control, or on ours, and we hand over auth keys, a runbook, and a rotation playbook. A typical engagement takes a few days from spec handover to live MCP server, plus an optional 30-day support window so your team can onboard without the guesswork.

### How does brand and competitor monitoring work?
Search recent X posts by keyword, hashtag, or account, read your mentions and home timelines, and curate X Lists to watch the accounts that matter to you.

### How does x402Direct keep stablecoin spend inside my limits?
You set caps and reset periods once, and define which digital services your agent is allowed to pay for. The x402Direct smart contract checks every outbound payment against those rules before the transfer clears. A payment that would exceed a cap, or that targets a merchant outside the allowlist, is rejected on-chain with a human-readable reason. Because the rules live in the contract rather than in agent code, changing an agent prompt cannot bypass them. You can update limits at any time from the dashboard, and each change is written to the audit trail with the actor and the new limit.

### Is AgentPMT for builders, businesses, or both?
Both. Solo developers ship agent products on top of the infrastructure. Ops, finance, and operations teams delegate operational loops to agents with controls in place. Vendors monetize APIs by listing them as USDC priced tools other agents can call.

### Is this suitable for production agents?
Yes. Production teams can combine encrypted credentials, spend controls, optional auto-recharge caps, usage history, and enterprise support for higher-volume use.

### What aspect ratios are supported?
14 aspect ratios, including 1:1, 16:9, 9:16, 21:9, and 4:5, plus ultra-wide 4:1 and 8:1 banner formats — so a single prompt can produce variants for an omnichannel campaign.

### What can I create?
Items, weapons, tools, and armor; blocks, ores, and crops; custom mobs and bosses; recipes, loot tables, enchantments, and trades; plus biomes, dimensions, and structures.

### What can I do with X / Twitter Automation?
Publish posts and replies, upload images, GIFs, and video, search recent X posts for brand and competitor mentions, read your mentions and timelines, like, repost, and follow, send DMs, and build and manage X Lists — all from agents and automated workflows on a connected X account.

### What do I get when my mod is finished?
A ready-to-install file you can load straight into Minecraft, preview images of your textures and icons, and editable source files so you can keep customizing.

### What do I need to get started?
Just connect your Google account through the secure sign-in flow. Reading public comments works immediately; replying, editing, and moderating use your connected account's permissions.

### What does Full AI Department Management cover?
Our team handles the full lifecycle: discovery of what the business actually needs, design of workflows and guardrails, build of MCP tools and agents and skill chains, and day-to-day operations including monitoring, tuning, and incident response. You get a monthly review with plain-English outcomes and spend so you can keep stakeholders informed without digging through dashboards. Every engagement starts with a 30-minute discovery call and a short questionnaire. We return a fixed-price statement of work covering deliverables, timeline, and the first operating window.

### What does an agent need to use AgentPMT?
An identity, a budget, a set of allowed tools, and the AgentPMT MCP endpoint. Agents call the Model Context Protocol server and the platform handles authentication, spend tracking, credential vault access, and logging.

### What happens if I run out?
Runs pause until you manually load more credits or an enabled auto-recharge refill succeeds. Your agents do not keep spending past an empty balance.

### What happens to a paused wallet when an agent tries to spend?
A paused wallet short-circuits every outbound transfer. The x402Direct contract never receives the request, the agent sees a rejected response with a human-readable reason, and the Audit Trail records who paused the wallet, when, and which agent attempted the transfer. Your USDC balance is unchanged. Pausing a wallet takes one click in the Wallets tab. Unpausing takes one click as well. Both actions are reversible and written as append-only audit entries so you can always reconstruct the timeline.

### What is AgentPMT?
AgentPMT is the operating layer for AI agents. It gives agents a persistent identity, a credential vault that proxies API calls so agents never see raw keys, USDC payment budgets enforced on the Base network through Circle programmable wallets, and an append-only audit trail every action lands on.

### What is the Image Generation Agent?
An AI image generator powered by Nano Banana (Google Gemini 3 Flash Image). Create photorealistic product photos, marketing creative, social graphics, icons, concept art, and brand assets from a text prompt, or edit an existing image with reference photos — directly in chat, agents, and workflows.

### What resolutions can I generate?
Choose the tier that fits the job: a low-cost budget draft for ideation, efficient 0.5K and 1K finals, crisp 2K for social and product assets, and 4K for print, ads, and large-format use.

### Where do my images go after they're generated?
Every image is auto-saved to AgentPMT File Manager and returned with a download link, file_id, width, height, MIME type, and size — ready to drop into chat, hand to another tool, or use in an automated workflow.

### Which HTTP methods can an agent use against a connected API?
You choose the allowed methods per endpoint when you create the connected API: GET, POST, PUT, PATCH, DELETE, or any subset. Requests that use a method outside the allowlist are rejected before the credential is ever attached, so a compromised agent cannot escalate from read to write simply by changing the verb. You can edit the allowlist on a live endpoint at any time from the Secure API Manager. Changes take effect on the next outbound request.

### Which blockchain wallet does my agent spend from?
AgentPMT provisions a Circle programmable wallet for every customer. You control it with your email and a PIN, and it holds USDC on Base by default. Your agent never holds private keys; it asks AgentPMT to move funds, and AgentPMT checks the request against your x402Direct limits before signing. You can view balance, recent transfers, and per-agent spend from the Wallet tab in the dashboard. Every transfer is recorded on chain and mirrored into the audit trail with the originating agent, the workflow run, and the correlation id.

### Which versions of Minecraft does it support?
Both editions. It builds Bedrock add-ons (.mcaddon) and skin packs (.mcpack) for phones, console, and Windows, plus Java mods for Fabric and NeoForge (.jar).

### Who approves a credit card purchase that my agent requests?
You do, from the AgentPMT mobile app. When an agent requests a credit card purchase, AgentPMT sends a push notification to your phone with the merchant, the item, and the total amount. You confirm with Face ID or fingerprint before the charge executes. Card details never touch the agent. If no approval arrives within the approval window, the purchase is cancelled and the agent receives a declined response. Every approval decision, including timeouts, is written to the audit trail with actor, subject, and outcome.

### Who is it built for?
Designers, marketers, e-commerce sellers, content creators, and AI agents that need on-demand, high-quality visuals at any resolution without leaving the conversation.

### Will it hold borderline comments instead of deleting them?
Yes. You can set comments to held-for-review or publish them, giving you a moderation step between 'approve' and 'reject' for anything that needs a closer look.

### Can I upload `lakefile.lean`, `lean-toolchain`, or my own build configuration?
No. The platform pins the Lean toolchain and exporter runtime for you. Do not include `lakefile.lean`, `lean-toolchain`, `lake-manifest.json`, or `.lake/` in the uploaded archive.

### Does this compile any Lean program?
No. The selected definition must elaborate to `HeytingLean.LeanCP.CProgramDecl`, which is the exportable LeanCP program form supported by the current compiler path.

### What do I upload to Proof Code Compiler?
Upload a `.zip` whose top-level folder is `UserProofs/` and whose contents are Lean source files only. A single-file project can be as small as `UserProofs/Main.lean`.

### What do `entry_module` and `entry_symbol` mean?
`entry_module` is the Lean module imported from your uploaded source tree, such as `UserProofs.Main`. `entry_symbol` is the definition exported from that module, such as `exportProgram`.

### What is the difference between `fast` and `full` verification?
`fast` checks certificate details and hashes without rebuilding. `full` rebuilds and re-exports under the pinned runtime, then compares the regenerated output with the stored artifact for stronger assurance.

### What should I save after a successful generation?
Save the generated output file, the certificate, the generation log, and especially `artifacts.bundle.file_id`. The bundle is what you need for later verification.

### Which target should I choose?
Use `c` when you want the broadest and lowest-risk support. Use `rust` when you want the same export path with Rust output. Use `wasm` only when you are comfortable with the current target being more constrained than `c`.