Agency exists when one person (the principal) manifests assent that another (the agent) act on the principal's behalf and subject to the principal's control, and the agent consents to act. The relationship is fiduciary from the moment it forms — no contract, no compensation, and no formality is required.

• Four elements: manifestation by P · action on P's behalf · P's right of control · A's consent.
• Control means a right to control, not day-to-day supervision. The right can be narrow and still support agency.
• Agency is a legal conclusion drawn from conduct — the parties cannot create or defeat it by saying so (§ 1.02).
• Ask this before anything else. Every downstream doctrine (authority, imputation, respondeat superior, loyalty) presupposes it.

Whether a relationship is agency depends on the elements in § 1.01, not on what the parties call it. A contract that disclaims agency, or calls someone an independent contractor, is evidence of the arrangement but does not decide the legal question.

• The most litigated proposition in the course. Ask what the relationship actually does, not what the paperwork says.
• Labels cut both ways: calling someone an 'agent' does not make them one either.
• Statutes may impose their own definitions for their own purposes (unemployment, wage law) — a separate inquiry entirely.

A manifestation is how assent or intention is communicated — words, writing, or any other conduct that conveys meaning to the person who observes it.

• Small section, load-bearing. Both actual and apparent authority are defined by whose manifestation reached whom.
• Conduct counts: putting someone behind a counter, giving them a title, letting them use your brand, tolerating past dealings.
• Silence or inaction can be a manifestation where the circumstances make speaking up the expected response.

The definitional section. It fixes the vocabulary used throughout: coagents, the three principal types, notice, person, power, and related terms.

• Disclosed principal — the third party knows an agent is acting for a principal and knows who the principal is.
• Unidentified (partially disclosed) principal — the third party knows there is a principal but not who.
• Undisclosed principal — the third party does not know a principal exists at all.
• Coagents share the same principal; a subagent (§ 3.15) is appointed by an agent to do what the agent owes the principal.
• Identify the principal type first: it determines who is a party to the contract (Ch. 6) and whether apparent authority is even available.

An agent acts with actual authority when, at the time of acting, the agent reasonably believes — based on the principal's manifestations to the agent — that the principal wants the agent to act that way.

• The perspective is the agent's, and the source of the belief is the principal.
• Measured at the moment of action, not with hindsight.
• Express authority is what was said; implied authority (§ 2.02) is what reasonably follows from it.
• Actual authority binds the principal even where the third party knew nothing about the arrangement — which is why it is the only route available for an undisclosed principal.

Actual authority covers what the principal designated or implied, plus acts reasonably necessary or incidental to accomplishing the principal's objectives, as a reasonable agent in that position would understand them in context.

• This is where 'implied authority' lives. Authority to manage a store implies authority to buy inventory; it does not imply authority to sell the store.
• Context includes the agent's role, past dealings, industry practice, and anything the agent has notice of.
• Authority to negotiate is not automatically authority to execute and bind — a distinction the casebook tests directly.

Apparent authority is the power to affect the principal's legal relations when a third party reasonably believes the actor is authorized and that belief is traceable to the principal's own manifestations.

• Perspective shifts to the third party; the source of the belief is still the principal.
• The agent's own assertions of authority can never, standing alone, create it. That is what defeats the claim in Hoddeson.
• Manifestations need not be individualized — signage, advertising, uniforms, titles, and a course of dealing can all reach the public.
• Two questions, kept apart: what did P communicate, and what authority would a reasonable person infer from that communication?
• Structurally unavailable against an undisclosed principal: a third party who does not know P exists cannot form a belief traceable to P.

An employer is liable for torts its employees commit within the scope of employment.

• Two independent findings are required: employee status and scope of employment. Both are unpacked in § 7.07.
• Every employee is an agent; not every agent is an employee. Non-employee agents do not trigger this rule.
• Liability here is vicarious and faultless — separate from the principal's own negligence under § 7.05.

Someone who made no manifestation of authority may still be liable to a third party who justifiably changed position in reliance on an appearance of agency — if that person intentionally or carelessly caused the belief, or knew of it and failed to take reasonable steps to correct it.

• The fallback when apparent authority fails for want of a manifestation by the principal.
• Requires detrimental reliance, which apparent authority does not.
• Grounded in the defendant's culpable act or omission rather than in any communication of authority.
• Overlaps with, but is distinct from, the principal's direct negligence under § 7.05.

An undisclosed principal is liable to a third party who justifiably changed position because of an agent acting on the principal's behalf without actual authority, if the principal knew of the conduct and its likely effect and did nothing to correct it. The principal also cannot rely on secret limits that cut the agent's authority below what such an agent would ordinarily have.

• The undisclosed-principal analogue of estoppel — apparent authority cannot operate here, so this rule fills the gap.
• Secret instructions are ineffective against a third party who reasonably assumed the ordinary scope of the agent's role.
• Pairs with § 6.03 on who is a party to the contract.

If a principal is unjustly enriched at another's expense through the acts of an agent or apparent agent, the principal may be liable in restitution.

• A safety valve when no attribution theory works but the principal has kept the benefit.
• Restitution measures the enrichment, not the plaintiff's expectation — a smaller recovery than contract.
• Distinguish from ratification: keeping a benefit can be evidence of ratification (§ 4.01) but is not the same claim.

Actual authority is created by the principal's manifestation to the agent, as the agent reasonably understands it, expressing assent that the agent act on the principal's behalf.

• Nothing needs to be said to the third party — this channel runs P → A only.
• The agent need not manifest assent back to the principal; consent can be shown by acting.
• Read with § 2.01 (what actual authority is) and § 2.02 (how far it reaches).

Where the law requires a signed writing to evidence an agent's authority to bind the principal, the principal is not bound without it. But a principal may be estopped from asserting the missing writing where a third party detrimentally relied on a reasonable belief in the agent's authority traceable to the principal.

• The equal-dignities rule: if the underlying transaction must be in writing, the authority to execute it generally must be too.
• Two different questions — did the agent actually have authority, and is there legally sufficient written evidence of it?
• A written designation may be too general to establish authority for the specific act (the majority/dissent split in Roger Gray).
• The estoppel clause matters: the formality is a defense the principal can forfeit by its own conduct.

Apparent authority arises from a manifestation by the person to be bound — made to the third party, directly or through others — that leads the third party reasonably to believe the actor is authorized.

• The manifestation may travel indirectly: through intermediaries, documents, position, or public display.
• Placing someone in a position that customarily carries certain powers is itself a manifestation.
• Trace the chain backwards from the third party's belief. If it dead-ends at the agent, the claim fails.

An individual can be a principal for any act the individual could lawfully do personally at that time. For entities, the law governing that entity determines capacity. Acts that cannot be delegated are not performed by using an agent.

• Minors, persons lacking capacity, and non-existent entities raise this issue.
• Nondelegable acts — voting, testifying, signing certain instruments — cannot be done through an agent at all.
• Contracting for an entity that does not yet exist pushes the liability onto the purported agent (§ 6.04).

Almost anyone can serve as an agent and affect another's legal relations. The agent's own capacity governs what duties and liabilities the agent personally incurs.

• A minor can bind a principal even where the minor's own contracts would be voidable.
• Capacity to act as agent and capacity to be personally liable are separate questions.

The roadmap section. Actual authority ends by: death or cessation of existence of either party; the principal's loss of capacity; agreement or changed circumstances; revocation by the principal or renunciation by the agent; or termination of the agency relationship itself.

• Sections 3.07–3.10 fill in each route; § 3.11 handles what survives.
• The exam point lives in what this section does not do — ending actual authority does not end apparent authority.
• Week 4's assigned cases are employment-at-will decisions, not applications of this section. Learn the section on its own.

An agent's death ends the agent's authority immediately. A principal's death also ends it, but only when the agent has notice — and it binds a third party only when that party has notice. Parallel rules apply when a non-individual agent or principal ceases to exist.

• The Third Restatement's notice requirement is a deliberate change from the older automatic-termination rule.
• Until notice arrives, an agent acting in good faith can still bind the estate.
• Powers given as security are carved out (§§ 3.12–3.13).

A principal's loss of capacity ends the agent's authority, effective when the agent has notice that the incapacity is permanent or has been adjudicated — and against a third party when that party has such notice. A written instrument may make authority effective on, or survive, incapacity.

• This is the doctrinal home of the durable power of attorney.
• Same notice architecture as § 3.07: incapacity alone does not silently strip authority.

Authority ends as the parties agreed, or when circumstances change enough that the agent should reasonably conclude the principal would no longer want the agent to act.

• Covers expiry dates, completed objectives, and events that make the assignment pointless.
• The standard is the agent's reasonable inference — no communication from the principal is required.
• Examples: the subject property burns down, the market collapses, war breaks out, the principal's instruction becomes illegal.

Either party can end actual authority unilaterally — the principal by revoking, the agent by renouncing — effective when the other has notice. This works even if it breaches their contract, though the breach may be actionable.

• Distinguish power from right: the principal always has the power to revoke, and may still owe damages for doing so.
• Exception: powers given as security and irrevocable proxies (§§ 3.12–3.13).
• Revocation is effective on notice — which is why the notice question in § 3.11 follows immediately.

Ending actual authority does not by itself end apparent authority. Apparent authority lasts until it is no longer reasonable for the third party to believe the agent still has actual authority.

• The most exam-worthy rule in Chapter 3, and no assigned case teaches it.
• Fire an agent Friday; a longstanding supplier who received no notice can still bind you Monday.
• The practical answer is notice — actual notice to those who dealt with the agent, and reasonable public notice to everyone else.
• The longer and more visible the prior course of dealing, the longer the appearance survives.

A power given as security is created in the form of authority but held for the benefit of the holder or a third person — to protect a title or interest or secure performance of a duty. An irrevocable proxy to vote securities is the same idea in the ownership-interest context.

• Not true agency: the holder exercises it for their own benefit, not the creator's, and owes no fiduciary duty as such.
• Requires an underlying interest or obligation. Calling a power 'irrevocable' without one does not make it so.
• Common in secured lending, escrow arrangements, and closely held corporation voting agreements.

Such a power ends when the secured obligation is discharged or the protected interest ends, when exercise becomes illegal or impossible, or when the beneficiary surrenders it. Ordinary revocation, renunciation, death, or incapacity does not end it.

• The mirror of § 3.10 — this is what 'irrevocable' actually means.
• The power dies with the underlying obligation, not with the person.

An agent acting for more than one principal in the same matter may be a subagent (§ 3.15), an agent for coprincipals (§ 3.16), or both.

• Classify before assigning duties: subagency and coagency route obligations differently.
• Serving two principals with adverse interests implicates §§ 8.03 and 8.06 (consent after full disclosure).
• Draw the chart — P, A, and the second actor — before answering anything.

A subagent is appointed by an agent to perform functions the agent has agreed to perform for the principal, and the appointing agent remains responsible to the principal for the subagent's conduct. Both relationships — subagent to agent, and subagent to principal — are agency relationships. An agent may appoint a subagent only with actual or apparent authority to do so.

• Test: is the second actor performing the first agent's obligations, or serving the principal directly as a coagent?
• The appointing agent stays on the hook — delegation transfers the task, not the responsibility.
• Even where delegation is authorized, the agent owes a separate duty of care in selecting and supervising the substitute (§ 8.08).
• Notice and knowledge can travel up a subagency chain (Ch. 5) — but only within each link's scope.

Two or more persons may jointly appoint one agent to act for all of them in the same matter.

• Each coprincipal is a principal in full — each can be bound, and each is owed the full fiduciary duty.
• Distinguish from subagency: the agent here answers directly to each principal, not through another agent.
• Conflicts among coprincipals trigger §§ 8.03 and 8.06.

Ratification is the affirmance of a prior unauthorized act, which then takes effect as if it had been authorized. A person ratifies by manifesting assent that the act count, or by conduct that reasonably implies that assent — provided the act is ratifiable, the ratifier has capacity, the ratification is timely, and it covers the whole act.

• A cure applied after the fact — it does not ask what anyone believed at the time of the transaction.
• Silence plus retention of a known benefit can be enough; the question is whether P made an informed choice.
• Four gates: ratifiable act (§ 4.03) · capacity (§ 4.04) · timing (§ 4.05) · entirety (§ 4.07). Knowledge is § 4.06.
• Run it last, after actual and apparent authority — never blended with either.

Ratification relates back and creates the effects of actual authority from the outset. It fails where it was induced by misrepresentation, where an agent invokes it against the principal to escape a loss, or where it would cut off rights third parties acquired in the meantime.

• Retroactivity is the whole point — the act is treated as authorized when it was done.
• The exceptions all protect someone who would be unfairly surprised by the relation-back.
• Ratification also ends the agent's exposure under the implied warranty of authority (§ 6.10).

An act can be ratified only if the actor purported to act as an agent on the ratifier's behalf.

• Someone acting purely for themselves leaves nothing to ratify — you cannot adopt a stranger's transaction.
• The actor need not have been an agent at all; purporting to act for the person is enough.
• Illegal acts and acts the ratifier could not have authorized are outside the section.

A person may ratify only if that person existed when the act was done and has capacity at the time of ratifying. A principal who ratified while lacking capacity may later avoid the ratification.

• The existence requirement is why a corporation cannot truly 'ratify' a promoter's pre-incorporation contract — it must adopt it as a new undertaking.
• Ties directly to § 6.04 and to promoter-liability problems in the entity weeks.

Ratification comes too late once it would unfairly affect third-party rights — after the third party withdraws, after a material change in circumstances makes it inequitable to hold them, or after a deadline that fixes rights.

• The third party can escape by withdrawing before the principal affirms.
• Classic example: a principal who waits to see whether the market moves before deciding to ratify cannot have it both ways.
• Insurance and option scenarios are the standard fact patterns.

A ratification made without knowledge of the material facts does not bind, unless the person knew their knowledge was incomplete and chose to affirm anyway.

• Informed choice is the core of the doctrine — not mere receipt of a benefit.
• Willful blindness does not help: consciously affirming despite known gaps is effective ratification.
• This is where Wing (no meaningful opportunity to reject) and Colony of Wellfleet (access to the facts, no timely repudiation) diverge.

Ratification must cover the entire act, contract, or transaction. You cannot adopt the favorable parts and disown the rest.

• Attempted partial affirmance is generally treated as no ratification — or as a counteroffer.
• Watch for principals who accept payment under a deal while contesting its burdens.

Someone who manifests that they have ratified, and thereby induces a third party to change position detrimentally, may be estopped from denying the ratification.

• Reaches conduct that falls short of true ratification but was reasonably read as approval.
• Same structure as §§ 2.05 and 3.02: detrimental reliance forfeits the defense.

A notification is a communication made in the required or a reasonable manner, intended to change legal rights and duties. A person has notice of a fact by knowing it, having reason to know it, receiving an effective notification, or having a duty that requires knowing it. Notice given to or by an agent can operate as notice to or by the principal.

• Distinguish notification (an act of communication) from notice (a state of awareness the law attributes).
• 'Reason to know' is an objective standard — actual awareness is not required.
• The chapter's two mechanisms follow: notification through an agent (§ 5.02) and imputation of what the agent knows (§ 5.03).

Notification to an agent is effective as notice to the principal if the agent has actual or apparent authority to receive it; notification by an agent counts as the principal's if the agent is authorized to give it. Not so where the person on the other side knows the agent is acting adversely to the principal.

• Authority is subject-matter specific: authority to sell is not authority to receive every legally significant communication.
• The principal cannot answer 'my agent never told me' — receipt by an authorized agent is the legally operative event.
• Not an evidentiary presumption about what the agent actually passed along. That is the point Farr makes.
• A third party who knows of the agent's adverse conduct cannot rely on this rule.

For determining the principal's relations with third parties, a fact the agent knows or has reason to know is imputed to the principal when that knowledge is material to the agent's duties — unless the agent is acting adversely (§ 5.04) or owes someone else a duty not to disclose it.

• Applies whether or not the agent ever told the principal, and whether the agent learned it by notification or independently.
• The materiality limit does the work: the fact must relate to what this agent was doing for this principal.
• Capacity matters — knowledge acquired while wearing a different hat is not organizational knowledge.
• The nondisclosure exception is why lawyers and other confidence-holders break the ordinary chain.

Notice is not imputed when the agent acts adversely in the matter, intending to serve only the agent's own purposes or another's. But it is imputed anyway to protect a good-faith third party, or where the principal ratified or knowingly kept the benefit. A third party who knows the agent is acting adversely is not in good faith.

• A narrow exception: the agent must effectively abandon the principal's interests, not merely have a conflict.
• Wrongdoing that still benefits the principal is not adversity — ask adverse to whom, and for whose benefit.
• The good-faith carve-out is why the exception protects innocent outsiders but not the agent's confederate.
• Retaining the benefit knowingly restores imputation — the principal cannot keep the fruit and disclaim the knowledge.

When an agent with actual or apparent authority contracts for a disclosed principal, the principal and the third party are the contracting parties; the agent is not, unless the agent and the third party agree otherwise.

• The ordinary case: the agent drops out of the transaction entirely.
• 'Unless they agree otherwise' is where personal guaranties and signature-block problems live.
• Sign in a representative capacity and identify the principal — that is the whole lesson of Flynn.

Where the principal is unidentified — the third party knows there is one but not who — both the principal and the agent are parties to the contract, unless the agent and third party agree otherwise.

• The default flips: the agent is liable, because the third party has no one else to look to.
• Naming the principal in the writing is the cheapest possible risk management.
• Trade names and shorthand designations frequently create this status by accident.

When an agent with actual authority contracts for an undisclosed principal, the agent and the third party are parties, and the principal is also a party unless the contract excludes it. Once a party, the principal has the same rights and liabilities as if it had contracted personally.

• Actual authority only — apparent authority cannot operate where the third party does not know P exists.
• The third party gets a windfall defendant it never bargained for, and the principal can enforce a contract it never signed.
• Setoff and settlement rules (§§ 6.06–6.08) protect the third party who dealt believing the agent was the principal.

Someone who contracts as agent for a principal that does not exist or lacks capacity becomes a party to the contract personally, if they know or have reason to know the problem, unless the third party agrees otherwise.

• Promoter liability. Sign for an LLC that has not been formed and you have signed for yourself.
• Later formation does not automatically cure it — the entity must adopt the obligation, and the promoter usually stays liable absent a novation.
• Runs straight into the entity weeks: this is the second holding in Philip Alan.

If the contract the agent made differs from the authorized one only in amount or by a separable part, the principal is bound to the extent of what was authorized — provided the third party promptly says it will accept that, and the principal has not already relied on there being no contract.

• Salvage rule for over-reaching agents: the deal is trimmed to the authorized size rather than voided.
• Requires severability plus a prompt manifestation from the third party.
• This is the doctrinal route to the result in Rowen — exposure capped at the authority actually granted.

Setoff rights depend on the principal's status. Where the principal is disclosed, the third party generally cannot set off debts the agent independently owes it. Where the principal is undisclosed, the third party may set off what the agent owed it up until the third party learns a principal exists.

• Protects a third party who extended credit believing the agent was the real party.
• The cutoff is notice of the principal's existence — everything accrued before it counts.
• Practical consequence of hiding a principal: you inherit the agent's credit history with that counterparty.

A principal's payment to its agent discharges the principal only if the principal reasonably relied on the third party's own manifestation that the agent had settled up. A third party's payment to the agent discharges the third party if the agent had actual or apparent authority to accept it. Special timing rules apply to undisclosed principals.

• Asymmetric on purpose: paying your own agent is generally your risk, not the third party's.
• Authority to collect payment is a separate authority question — many agents can sell but not receive funds.
• Undisclosed principals bear the risk of the third party settling with the agent before learning of them.

Later dealings between the agent and the third party can enlarge or reduce the principal's rights and liabilities where the agent acts with actual or apparent authority or the principal ratifies. For an undisclosed principal, dealings before the third party learns of the principal generally bind it.

• Modifications, releases, and waivers run through the same authority analysis as the original contract.
• Another cost of concealment: the undisclosed principal is bound by a course of dealing it never participated in.

Obtaining a judgment against either the agent or the principal does not discharge the other's liability. It is discharged only to the extent the judgment is actually satisfied.

• The Third Restatement rejects the old election-of-remedies rule: one recovery, but not a forced choice of defendant.
• Williams v. Investors Syndicate applies the older Massachusetts election doctrine and comes out the other way — an excellent 'was this correctly decided?' comparison.
• Practical point: sue both, satisfy once.

Someone who purports to act for another without the power to bind them impliedly warrants their authority and is liable for the loss that breach causes — including the benefit expected from the principal's performance. No liability if the principal ratifies, if the agent disclaimed the warranty, or if the third party knew there was no authority.

• The third party's remedy when authority fails entirely: sue the agent on the warranty.
• Liability is not for the contract itself but for breach of the warranty — though damages can reach expectation.
• Good faith is no defense; an honest agent who overestimated their authority still warrants it.

A false statement by the agent about the agent's own authority binds the principal only where the agent had actual or apparent authority to make it and the third party lacks notice of its falsity. Representations made incident to an authorized contract are attributed to a disclosed or unidentified principal as if made by the principal.

• Guards against bootstrapping — the agent's claim of authority cannot manufacture the authority.
• But representations about the subject matter of an authorized deal do bind the principal.
• A third party with notice of falsity cannot rely.

An agent who commits a tort is personally liable to the person harmed. Acting with authority, or within the scope of employment, does not shield the agent.

• Vicarious liability adds a defendant; it never subtracts one.
• Employees are personally liable for their own torts even when the employer must also answer.
• Statutes occasionally provide immunity — public employees under a tort claims act, for example.

An agent's breach of duty owed to the principal is not, by itself, a basis for liability to a third party. The agent is liable to the third party only for breaching a duty owed to that third party.

• Keeps the internal relationship out of the outsider's claim.
• The third party must find an independent duty — negligence, misrepresentation, a statute.
• The mirror image of § 7.01: separate duties, separate claims.

The organizing section. A principal is directly liable for authorized or ratified tortious conduct, for its own negligence in selecting or controlling an agent, and for failing to perform a nondelegable duty. A principal is vicariously liable for an employee's torts within the scope of employment and for torts committed with apparent authority.

• Name the theory before you argue it. Direct and vicarious liability require entirely different proof.
• Direct: §§ 7.04, 7.05, 7.06. Vicarious: §§ 7.07, 7.08.
• Direct liability reaches non-employee agents and independent contractors, where respondeat superior cannot.

A principal is liable for harm caused by an agent acting within actual authority or by conduct the principal ratified, where the conduct is tortious or would be tortious if the principal had done it.

• Direct, not vicarious — the principal effectively ordered the conduct.
• Applies to non-employee agents too: employee status is irrelevant here.
• Covers the principal who instructs an agent to make representations that turn out to be false.

A principal is liable for harm caused by an agent where the principal was negligent in selecting, training, retaining, supervising, or controlling that agent. Where the principal has a special relationship with the person harmed, it owes that person reasonable care against risks arising from the relationship, including risks posed by its own agents.

• The principal's own fault — available even where the actor is an independent contractor.
• The theory in Tormo (negligent selection of a substitute) and the theory attempted in Cullen (negligent supervision of a dealer).
• Requires a duty, a breach, and causation like any negligence claim — the failed element in Cullen was duty.
• Special relationships: schools, carriers, innkeepers, custodial institutions.

A principal who owes a duty of protection by contract or by law cannot escape it by delegating performance, whether or not the delegate is an agent.

• Nondelegable duties: the work can be handed off, the liability cannot.
• Common in premises safety, common-carrier obligations, and statutory duties.
• Defeats the 'we hired a contractor' defense where the duty runs to the plaintiff directly.

An employer is vicariously liable for an employee's torts within the scope of employment. Conduct is within scope when the employee performs assigned work or engages in a course of conduct subject to the employer's control; it is outside scope when it occurs in an independent course of conduct not intended to serve any employer purpose. An employee is an agent whose principal controls or may control the manner and means of the work — and gratuitous work still counts.

• Two findings, in order: employee status (control over manner and means), then scope.
• The Third Restatement replaces 'frolic and detour' with a functional test: an independent course of conduct serving no employer purpose.
• Intentional torts can fall inside scope where they arise from the work itself.
• Statutory definitions of 'employee' are separate inquiries and can come out differently on identical facts.

A principal is vicariously liable for an agent's tort committed while dealing or communicating with a third party on the principal's behalf, where the agent's apparent authority either constitutes the tort or enables the agent to conceal it.

• Extends vicarious liability beyond employees to anyone cloaked with apparent authority.
• Built for misrepresentation, fraud, and abuse-of-position cases — the appearance is what made the harm possible.
• The tort-side companion to § 2.03, and the doctrinal home of apparent-agency claims against franchisors.

An agent owes a fiduciary duty to act loyally for the principal's benefit in everything connected with the agency.

• The umbrella. Sections 8.02–8.05 are specific applications, not an exhaustive list.
• Applies to gratuitous agents and to relationships neither party labeled fiduciary.
• Remedies are restitutionary as well as compensatory: disgorgement, constructive trust, forfeiture of compensation.
• Partners are agents of one another — which is how this chapter reappears throughout the partnership weeks.

An agent may not take a material benefit from a third party in connection with transactions done for the principal or otherwise through use of the agent's position.

• Kickbacks, referral fees, side payments, and undisclosed commissions.
• No harm to the principal is required — the benefit itself is the breach.
• Curable only by consent meeting § 8.06's disclosure conditions.
• Remedy: the agent disgorges the benefit to the principal.

An agent may not deal with the principal as, or on behalf of, an adverse party in a transaction connected with the agency.

• Self-dealing and dual representation — the agent on both sides of the table.
• Applies to the agent's own purchase from or sale to the principal.
• Only § 8.06 consent saves it, and that requires disclosure of everything material.
• The general-partner and LLC-manager cases run on exactly this principle.

Throughout the agency, the agent must not compete with the principal or assist the principal's competitors. During that same period, the agent may take steps that are not otherwise wrongful to prepare to compete after the relationship ends.

• Two rules in one sentence. The line between preparation and competition is the whole litigation.
• Generally permissible while employed: planning, incorporating, arranging financing, leasing space, buying equipment.
• Generally not: soliciting the principal's customers or employees, diverting opportunities, using confidential information.
• The duty ends with the agency. Post-termination restraints require a valid contract or statute — not this section.
• Confidentiality (§ 8.05) survives independently, which is why 'free to compete' never means 'free to use the files.'

An agent must not use the principal's property, or use or disclose the principal's confidential information, for the agent's own purposes or anyone else's.

• The duty regarding confidential information outlives the agency — unlike the duty not to compete.
• Covers customer lists, pricing, methods, and internal data whether or not they qualify as trade secrets.
• General skill and knowledge the agent acquired on the job remain the agent's own.
• This is the doctrinal difference between competing and competing unfairly.

Conduct that would breach §§ 8.03–8.05 is not a breach if the principal consents — but only where the agent acted in good faith, disclosed all material facts that would reasonably affect the principal's judgment, otherwise dealt fairly, and the consent covers the specific act or a defined type of act.

• Consent is a procedure, not a formality. Blanket advance waivers rarely satisfy it.
• Disclosure must precede consent and must include what the agent knows, has reason to know, or should know.
• Silence, acquiescence after the fact, or consent given in the wrong manner does not count — the lesson of Bassan.
• A sufficiently specific agreement can authorize what would otherwise be usurpation, which is how Fronk comes out for the general partners.

An agent must act in accordance with the express and implied terms of any contract with the principal.

• Contract duties sit alongside fiduciary duties; breaching one is not automatically breaching the other.
• Different remedies follow: expectation damages for contract, disgorgement and forfeiture for disloyalty.

An agent must act with the care, competence, and diligence normally shown by agents in similar circumstances. Special skills the agent has — or claims to have — raise the standard accordingly.

• Negligence, measured against the relevant class of agents, and modifiable by agreement.
• Claiming expertise you lack imports the expert's standard.
• Includes the duty to use care in selecting and supervising a subagent, which is the theory in Tormo.
• Professional malpractice is this duty applied to lawyers, brokers, and other skilled agents.

An agent must act only within actual authority and must comply with the principal's lawful instructions.

• The internal counterpart to Chapter 2: exceeding authority can bind the principal externally and still breach this duty internally.
• 'Lawful' does the limiting work — an agent need not follow instructions to commit a wrong (see § 8.10).
• The agent who exceeds authority also risks the implied warranty of authority (§ 6.10).

Within the scope of the relationship, an agent must act reasonably and avoid conduct likely to damage the principal's enterprise.

• Reaches conduct that harms the principal's reputation or operations without being disloyal in the § 8.04 sense.
• Supplies the agency-law answer to why an agent may refuse an instruction to act unlawfully.

An agent must use reasonable effort to give the principal facts the agent knows, has reason to know, or should know — where the principal would want them or they are material to the agent's duties — unless a superior duty to someone else forbids disclosure.

• The affirmative disclosure duty. Concealment is a breach even without self-dealing.
• The superior-duty limit is the same carve-out that blocks imputation under § 5.03(b).
• Cardozo's disclose-and-offer answer in Meinhard is this duty applied to a joint venture.

An agent must not treat the principal's property as the agent's own, must not commingle it with anyone else's, and must keep and render accounts of money and property received or paid on the principal's behalf.

• Segregation, record-keeping, and accounting — the trust-account rules for lawyers and brokers in general form.
• Commingling shifts the burden of proof onto the agent to establish what belongs to whom.
• Reappears as the intermingling factor in veil-piercing analysis.

A principal must act in accordance with the express and implied terms of any contract with the agent.

• The principal's contract duty — compensation, expenses, and the agreed terms of engagement.
• The power to terminate (§ 3.10) does not include the right to breach without consequence.

A principal must indemnify the agent as the contract provides, and otherwise for payments made within actual authority or beneficially (unless officious), and for losses that fairly should fall on the principal given the relationship.

• Covers expenses advanced, liabilities incurred, and, in many cases, defense costs.
• Fault matters: an agent who breaches a duty or knowingly acts wrongfully generally loses the right.
• The corporate-indemnification statutes are the codified version of this principle.

A principal must deal fairly and in good faith with the agent, including warning the agent about risks of physical harm or financial loss in the work that the principal knows or should know about and the agent does not.

• The reciprocal duty — the relationship is fiduciary in one direction and good-faith in the other.
• The warning obligation covers hazards, undisclosed liabilities, and known defects in the assignment.
• The employment-at-will cases test how far this good-faith idea reaches into termination, and jurisdictions split sharply.