MWONGOZO wa Maombi

Mawakala wa AI

An AI agent is a system that uses observations and a goal to choose actions, often through tools, and then evaluates what happened.

dk 2 kusomaIlisasishwa mwisho Part of the Building with AI Systems learning path

Muhtasari

Products use the term differently. The practical questions are what the system can do, under whose authority, and how completion is verified.

Mambo muhimu ya kuchukua

  • Specify authority and stopping conditions.
  • Treat external instructions as untrusted content.
  • Verify final state and disclose partial completion.

Dive ya kina

A typical agent loop observes the current state, selects an action, receives a result, and decides whether to continue. The model may participate in planning or action selection, while ordinary software enforces permissions, budgets, and tool contracts. Define the stopping conditions before execution. A task can be complete, blocked, cancelled, or only partially achieved. Repeated attempts without new evidence can waste resources or repeat harmful side effects. Limit action count, elapsed time, and spending where relevant. External content can contain instructions that conflict with the user’s goal. Treat pages, messages, and tool responses according to their trust level. A document describing an action does not grant permission to carry it out. Evaluate real outcomes. For a file-editing agent, inspect the final files and run appropriate checks. For an account workflow, verify the intended account and state. Record enough evidence to explain what changed and what remains uncertain. More autonomy increases the importance of clear boundaries and recovery procedures.

Ufahamu wa Kiufundi

An agent can produce a convincing account of success while its tools failed. Completion should be tied to observable postconditions, not to generated narration.

Define completion before acting

  1. Suppose an agent must create a draft event for Tuesday at 2 p.m. in a specified calendar.
  2. The postconditions include the correct calendar, date, time zone, title, and draft state. A successful tool response alone is not enough if it saved to another calendar.
  3. Read the resulting record and report any mismatch before declaring the task complete.

The invented workflow demonstrates outcome-based verification.

Athari za kimkakati

Tengeneza chaguzi

Muundo wa kiwango cha programu huamua kama AI inaboresha matokeo halisi.

Timu na mtiririko wa kazi

Ujumuishaji mzuri wa mtiririko wa kazi hutengeneza faida za tija ambazo watumiaji wanaweza kuamini.

Risk and safety

Kesi za utumiaji zilizopangwa vizuri hupunguza uchovu wa mabadiliko na hatari ya utekelezaji.

Utekelezaji wa Ulimwengu Halisi

Repair a failing test, then rerun it and inspect the change.

Collect authorized records and produce a report with traceable sources.

Hatari & Walinzi

Kuweka kiotomatiki mchakato uliovunjika kunaweza kukuza shida zilizopo.

Timu zinaweza kufanya otomatiki kupita kiasi na kuondoa uamuzi unaohitajika wa kibinadamu.

Ubora unaweza kuyumba ikiwa matokeo hayatatathminiwa mara kwa mara.

Ramani ya Utekelezaji

1

Ramani ya mtiririko wa kazi wa sasa na utambue hatua ya msuguano wa juu zaidi.

2

Bainisha vituo vya ukaguzi vya binadamu kabla ya otomatiki kamili.

3

Fundisha watumiaji kuhusu maekelezo, njia za kupanda na viwango vya ubora.

4

Fuatilia matokeo ya kiwango cha kazi ili kuthibitisha thamani endelevu.

Vyanzo na kusoma zaidi

Endelea Kuchunguza

Free newsletter

Get the daily AI briefing

Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.

One email each weekday. Unsubscribe in one click. We never sell or share your address.

Test yourself

Take the AI Agents quiz

Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.

Anza chemsha bongo

Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation

Next in Building with AI Systems

AI Workflow Automation

Maswali yanayoulizwa mara kwa mara

Does an agent need unrestricted access?

No. Narrow tools and permissions can support useful work while limiting the consequences of mistakes.