# AI agents for freight forwarders > Describe an agent in plain words and Fretie builds it: quote chasing, demurrage and detention watch, carrier notices, exception triage. Scoped permissions, suggest-first mode, full trace of every run. Agent Studio lets a forwarder build their own AI agents without writing code. Describe the job in plain words, or start from a preset, and Fretie drafts the agent: a name, instructions, the permissions it needs and a schedule. You review, adjust, and switch it on. Agents act through the same API your team uses, only within the permissions you tick. Every change can land as a suggestion for a person to apply, every run keeps a full trace, and no agent ever sends an email on its own. ## Describe it, Fretie builds it Type what you want in the language you would use with a new colleague: "Every morning at 8, find quotes sent more than 3 days ago with no reply and draft a friendly follow-up." Fretie turns that into a named agent with instructions, the permissions it needs, and a schedule. Or start from a preset: Quote Chaser, D&D Guardian, Notice Applier, Exception Triage, Inbox Responder, Invoice Auditor. ## Permissions per area, per operation Each agent gets a permission grid: quotes, inbox, rates, shipments, customers, reconciliation, exceptions, automations and templates, each with read, create, update and delete. Anything not ticked is refused at the API, even if the instructions ask for it. An agent that can read quotes and create drafts cannot touch your rate cards. ## Suggest first, or fully automated In suggest-first mode, every change lands as a suggestion with an Apply button and nothing happens until a person applies it. In fully automated mode changes apply immediately, behind a confidence gate: any write the agent is less sure about than your threshold becomes a suggestion instead. Both can be overridden per area. ## Runs on a schedule, on events, or on demand Run every few hours, daily, or weekly. Or react to events: a quote created, a shipment status change, a carrier notice recorded, an invoice received, an exception raised. Each agent has a daily run budget and a step limit, so a runaway loop cannot become a runaway bill. ## It learns from your team A team member can opt in to teach one agent. The dashboard records what they do (pages opened, clicks with the control's label, values typed, never passwords or secret fields) and the agent distils procedures, preferences and patterns it then follows. Learnings can be dismissed, deleted, or promoted to a reusable skill shared across agents. ## Every run traced, every credit accounted for History shows every run with each tool call, each API request and the credits it used. Every action is in the audit log. A daily digest tells you what your agents did, and you get an alert if one fails twice in a row. Chat with any agent to ask what it did and why. ## What you get - Plain-language builder: name, instructions, permissions and schedule drafted for you - Six presets: Quote Chaser, D&D Guardian, Notice Applier, Exception Triage, Inbox Responder, Invoice Auditor - Permission grid per area and per operation, enforced at the API - Suggest-first mode: nothing changes until a person clicks Apply - Fully automated mode with a confidence gate and per-area overrides - Schedules, event triggers and on-demand runs, with a daily budget and step limit - Full trace per run: tool calls, API requests and credits used - Skills hub: procedures written by hand or promoted from what an agent learned - Memory between runs and dismissal reasons the agent reads before it starts - Drafts go to an outbox; a person sends every email - Daily digest, and an alert when an agent fails twice in a row ## Frequently asked questions ### Can an agent send email to my customers on its own? No. Agents never send email. Anything an agent writes lands in an outbox as a draft, and a person on your team sends it. That holds in fully automated mode too. ### What can an agent access? Only what you tick. Permissions are set per area (quotes, inbox, rates, shipments, customers, reconciliation, exceptions, automations, templates) and per operation (read, create, update, delete). The check happens at the API, so an instruction cannot talk an agent past its permissions. ### Can I see what an agent did? Yes. Every run has a full trace in History: each tool call, each API request and the credits it used. Every action is also in the audit log, and you can chat with the agent to ask what it did and why. ### What does it cost to run agents? Each run shows the credits it used, and each agent has a daily run budget and a step limit you set. A run on a stronger model costs more credits, shown before you pick it. Each run and each learning pass spends credits from your plan, the same credits that pay for quotes and document extraction, so there is nothing extra to subscribe to. See pricing for the credits included in each tier. ### How does learning respect privacy? Learning is opt-in per person and per agent. The recording captures page opens, clicks with the control's label and typed values, never passwords or secret fields. Everything is scoped to your workspace, never used to train shared models, and can be deleted in one click. ### Can I stop an agent? Yes. Pause or disable any agent at any time, switch it to suggest-first mode so nothing applies without a person, or tighten its permissions. Dismissing a suggestion with a reason also teaches the agent, and it reads those dismissals before every run. --- Source: https://fretie.com/features/ai-agents