Definition
AI Agent
An AI agent is a software system that uses a large language model to pursue a goal across multiple steps — reading context, choosing tools, and taking actions — rather than answering a single prompt. In a work platform, agents triage issues, draft updates, and execute multi-step tasks as autonomous teammates, bounded by the permissions and approvals their operators set.
Key takeaways
- An AI agent uses a large language model to pursue a goal across multiple steps — reading context, choosing tools, and taking actions — not just answering one prompt.
- The difference from a chatbot is autonomy over a loop: observe, decide, call a tool, read the result, repeat until the goal is met.
- Tools with typed definitions let the agent take real actions (triage an issue, draft an update), bounded by the permissions its operator sets.
- Autonomy is the risk — destructive actions should be gated behind approval and recorded in an audit trail so the agent stays accountable.
A chatbot responds; an agent acts. The distinction is autonomy over a loop: an agent is given an objective, then repeatedly observes its environment, decides a next step, calls a tool, and reads the result — continuing until the goal is met or it hits a stop condition. The model is the reasoning engine, but the tools (search, create an issue, post an update) are what let it change the world rather than just describe it.
What makes an agent useful in a work platform is that the same tools a human uses become callable by the model under explicit, typed definitions. An agent can read the backlog, summarize a stalled project, propose a re-prioritization, or draft a customer reply — chaining several real operations into one outcome instead of leaving the human to click through each step.
Autonomy is also the risk. An agent that can delete, archive, or bulk-edit data needs its reach bounded by policy, not just by the model's judgment. The mature pattern is to separate read-only actions, which can run freely, from destructive ones, which are emitted as proposals a human or guardrail confirms — and to record every action in an audit trail so the agent's behavior is reviewable.
Planoda puts an AI agent on every surface and runs destructive actions through a propose/approve guardrail, with each proposal and approval written to the same audit trail as human edits — so agents act as accountable teammates, not unsupervised automation.
Related terms
- Propose / Approve (AI Governance)Propose/approve is a governance pattern for autonomous software: instead of executing a consequential action directly, an AI agent emits it as a proposal that a human or policy must approve before it runs. It keeps fast, read-only work autonomous while gating destructive or irreversible operations — the practical way to give agents real power without surrendering control.
- MCP ServerAn MCP server implements the Model Context Protocol, an open standard that lets AI assistants connect to external tools and data through a uniform interface. Rather than building a bespoke integration per assistant, a tool exposes one MCP server describing its available actions and resources, and any MCP-capable AI can discover and use them safely.
- Semantic SearchSemantic search finds results by meaning rather than exact keywords, using vector embeddings that place similar concepts near each other in mathematical space. A query for 'login broken' can surface an issue titled 'users can't authenticate' even with no shared words. It powers more relevant search and is the retrieval layer behind many AI features.
- Audit TrailAn audit trail is an append-only, time-ordered record of who did what, when, and to which object across a system. Every create, edit, delete, and approval is logged immutably, so any state can be traced back to the actions that produced it. Audit trails underpin accountability, debugging, compliance, and — increasingly — oversight of what AI agents do.
- TriageTriage is the process of reviewing newly reported issues and deciding what happens to each — accept and prioritize, request more detail, route to a team, or close. Borrowed from emergency medicine, it keeps the incoming flow of bugs and requests from overwhelming a team by quickly sorting signal from noise at the front door.