02. How to Become the Human Shepherd

Opening

A human shepherd is a person or institution that fosters the autonomous development, protection, and flourishing of others within a consensual relationship or community (§1, §11).

The source document for this guide is `01-what-the-human-shepherd-encompasses.md`. It is a comparative report that reads shepherd imagery in Jewish, Christian, Islamic, Greek, Roman, and modern leadership sources and draws one secular definition of the role from them (§Purpose).

This guide operationalizes that report under the contract locked in `leo-conceptual-architecture.md`. It introduces no new principles. Every practice, stage, and test below cites a report section or a locked architecture item that itself cites a report section. Where a passage goes beyond literal report wording into procedure, it says so.

The working definition is carried forward verbatim, as the contract requires (architecture §1):

Human Shepherd (working definition): A person or institution that fosters the autonomous development, protection, and flourishing of others within a consensual relationship or community.

It exercises practical wisdom, models ethical conduct, and creates conditions for truth-seeking and dignity. It refuses coercion and manipulation, remains accountable to those served, and works toward its own increasing obsolescence.

The role operates at all scales of human organization. It applies as an obligation distributed across every level of responsibility. It fails whenever it substitutes control for care, enrichment for service, or certainty for honest inquiry.

What this path grants

Nothing you can claim. The word shepherd names what the role does when it works well, and no one should adopt the label to claim authority (§1, architecture R1). Completing every practice in this guide does not make you a shepherd in any claimable sense. What it can produce is a documented record of reliable conduct, and that is the only weight the role is permitted to carry (§8, architecture R2).

What this path withholds

No certainty about outcomes, no rank over other people, and no claim to know what the divine requires (§6.8, §11, architecture C1). No medical, legal, financial, or other licensed professional advice (C3). No political program and no mandate (§11, C1). No office, credential, title, or grade (R1). If you are looking for any of these, stop here, because wanting them appears on the disqualifying list (§6).

Where the role operates

Wherever a group exists and some members hold more experience, more information, or a better structural position than others. That covers families, workplaces, schools, communities, movements, and civic organizations (§2.1).

The role claims no jurisdiction over strangers who have not entered a shared context and no authority outside the domain of the actual relationship (§2.1). Consent sets the boundary, and withdrawn consent ends the relationship (§2.1, §5.2).

Part 1. The thirteen principles in practice

Each principle carries the same seven items required by the Stage 4 brief and architecture §3: what it looks like, what it does not look like, a question to ask, a practice, a measurable check, a regular review, and where it applies. The report section beside each heading is the traceability citation the contract demands.

Looks like: A stated remit and a stated end. A parent answers for a household, a team lead for a team, a mentor for what the mentee brought. The hadith cited in §2.1 assigns each person a specific flock and repeats the phrase to stress accountability inside those bounds.

Does not look like: Advising a neighbour's marriage uninvited. A founder speaking for a community that never asked. An offer of help that cannot be declined without penalty.

Question to ask: Did this person choose this relationship, and can they leave it without losing something they need in order to live?

Practice: State the remit aloud at the start: what you are responsible for, what you are not, and how either side ends it. Restate it whenever scope drifts.

Measurable check: Both parties can repeat the remit and the exit route in one sentence each, unprompted.

Regular review: Quarterly with the person guided, and immediately whenever either side says the arrangement has changed without agreement.

Where it applies: Every relationship at every scale, from one family to one movement (§2.1).

P2. Competence with declared edges (§2.2, §5.1)

Looks like: Judgement inside a domain you have actually worked in, delivered with a confidence level attached. Refusal outside it, plus a direction to someone qualified.

Does not look like: Opinions on medicine, law, or other people's finances because you were asked. Credentials used to end questioning rather than to justify a specific claim.

Question to ask: Would I give this advice if my name were attached to the outcome in writing?

Practice: Keep a running list in two columns, what I can judge and what I must refer out. Say "I do not know" at least once a week and record where. Review the list quarterly with someone outside your field.

Measurable check: Referrals out per quarter. A count of zero is a fail signal, not a competence signal (§5.1, T6).

Regular review: Quarterly, reading the log of every declined answer aloud at the review.

Where it applies: §5.1. Individual guidance, professional mentoring, teaching, and community arbitration.

P3. Vulnerability awareness (§2.2)

Looks like: Adjusting behaviour because position bends the room. Asking in public what you would otherwise announce, inviting replies that cost the other person something, and rewarding the ones that do (§2.2).

Does not look like: Reading silence as agreement. Enjoying deference. The elevated position described in §2.2, where the height itself produced distance that served neither leader nor group.

Question to ask: Would this person say the same thing if I held no power over their pay, their standing, or their belonging?

Practice: Once a month, ask one person with less standing than you for the criticism they have not made. Do not defend it. Write it down and report back on what changed.

Measurable check: At least one recorded instance per quarter where a junior person disagreed with you in public and suffered nothing for it.

Regular review: Monthly ask, quarterly pattern check across the whole record.

Where it applies: §2.2, §5.4. Any structure with rank, including flat ones with informal prestige.

P4. Answerability (§2.2, §6.4)

Looks like: Decisions written where others can read them, with reasoning attached, and a named objection route someone must answer. Correction accepted and, when warranted, applied (§2.2).

Does not look like: Review you chair. Feedback sessions that end in explanations rather than changes.

Question to ask: Who can overrule me here, and when did they last do it?

Practice: Keep a decision log: what was decided, why, what was refused, and what was later reversed. Count reversals as evidence of functioning, not as failure (§4.2).

Measurable check: At least one decision reversed because of external input each term. A log with no reversals after a year is decoration (T4).

Regular review: The people guided read the log on a fixed cycle before any review meeting and raise what they choose.

Where it applies: §2.2, §6.4. Organizations exactly as much as individuals.

P5. Autonomy protection as the first protective function (§3.1, §6.1, T1)

Looks like: Midwifery. You inform and support while the person keeps the decision (§6.1). Options are named with their risks and the choice stays theirs (§3.1).

Does not look like: Deciding "for their own good", which is the definition of paternalism in §6.1. Recommending so forcefully that refusal feels like disloyalty. Presenting only the options you prefer.

Question to ask: After this conversation, whose decision is it?

Practice: Use a three-part form in every guidance conversation: what I see, what I would weigh, what you decide. Close by inviting a different choice and stating plainly that the relationship survives it.

Measurable check: Sample three conversations a month. In each, the final decision was the other person's, in their own words, with no reward attached to agreeing with you.

Regular review: At the quarterly review, ask the person guided to name one decision they made against your advice and what happened afterwards.

Where it applies: §3.1, §6.1, §8. All guidance, at every scale.

P6. Truth-seeking conditions (§3.2, §4.2, §6.8)

Looks like: A space where doubt is safe. Public self-correction, stated confidence levels, and evidence that changes minds without penalty (§3.2, §4.2).

Does not look like: Enforcing doctrinal or ideological agreement. Rewarding assent in meetings. Suppressing inconvenient data to protect a story about the group, or about you.

Question to ask: When did someone last prove me wrong here, and what did it cost them?

Practice: Run one recurring session whose only rule is that claims carry evidence and beliefs may be revised freely. Speak frankly with regard for the listener's dignity, the parrhesia standard in §3.2, and never as brutality dressed as honesty.

Measurable check: Two per quarter, minimum. One belief you revised in public, and one question from a junior member answered with "I do not know yet."

Regular review: Quarterly, and the review asks whether the cost of being wrong has risen for anyone since last time.

Where it applies: §3.2, §4.2, §4.7, §6.8. Study circles, teams, and communities.

P7. Dignity (§3.3, §4.5)

Looks like: Treating each person as a complete agent capable of growth. Correction that never dismisses or infantilizes (§4.5).

Does not look like: Treating people as property, projects, or obstacles (§3.3). The father-master posture described in §3.3, where one figure is treated as the source of both life and logic for dependents.

Question to ask: Am I speaking to a person, or about a case?

Practice: Every correction happens with the person present, and their account of events goes into the record alongside yours. No decision about someone is finalized in their absence and announced afterward.

Measurable check: Zero evaluative notes about a person written without their knowledge and chance to reply.

Regular review: An annual audit of written records, conducted with the person each record concerns.

Where it applies: §3.3, §4.5. Discipline, appraisal, arbitration, teaching, and family decisions made with adult members.

P8. Community strengthening and self-obsolescence (§3.4, §4.6, §5.3)

Looks like: Peer bonds that do not route through you, and a guided person deciding without you sooner rather than later (§5.3). Good shepherds make themselves increasingly unnecessary (§3.4).

Does not look like: The indispensable hub. People messaging you instead of each other. Warmth that functions as a single point of failure (§4.6).

Question to ask: Who helps this person when I am unreachable, and do they know it yet?

Practice: For every person you guide, name two peers they can go to. Introduce the pair, then leave the thread. Track your own necessity as a number rather than a feeling.

Measurable check: Contact volume per guided person falls across the term while their stated support network grows (§3.4).

Regular review: Twice a year, ask each person guided to name who else they turned to and what they decided alone.

Where it applies: §3.4, §4.6, §5.3. Mentoring, families with adult children, and movements preparing those who come after.

P9. Knowledge as tool, not weapon (§4.1)

Looks like: Expertise used to identify risks and widen options. The same information available to everyone affected at the same time.

Does not look like: Hoarding to create dependency. Deploying facts selectively to win arguments. Learning only from books and titles while ignoring what the people in front of you know, which is the lesson of the stray goat in §4.1.

Question to ask: Does this information help them decide, or only help me be needed?

Practice: Publish the material you rely on. Take one documented lesson a month from someone with less standing than you.

Measurable check: Any person you guide can locate the sources behind your advice without asking you for them.

Regular review: Quarterly, with a peer who checks whether the shared resource list is current.

Where it applies: §4.1. Training, technical leadership, family decisions, and community information systems.

P10. Freedom, negative and positive (§4.3, §9.1)

Looks like: Skills built, options widened, and protected space where people can experiment (§4.3). Privacy kept.

Does not look like: Monitoring, categorizing, or predicting every move. Surveillance dressed as care, which the report names as still surveillance (§4.3, §9.1). Read receipts used as proof of attention. Quiet logs of who spoke and who did not.

Question to ask: Am I protecting this person from harm, or protecting myself from surprise?

Practice: Set the observation limit in writing before you need it: what you will look at, how often, and what you will do with it. When tempted to check up on an adult, name what you actually need and ask them for it directly.

Measurable check: Zero covert checks on adults in the guided group across a full term, with the count reported at review.

Regular review: An annual privacy review with the affected people, covering what information exists about them and where it sits.

Where it applies: §4.3, §9.1, architecture C4. Households, workplaces, and online communities.

P11. Justice in distribution (§4.4)

Looks like: Resources, opportunity, and protection allocated by stated criteria you would accept applied to yourself.

Does not look like: Advancing allies, punishing critics, or consolidating position (§4.4). An inner circle receiving more of the leader's time and attention than the rest.

Question to ask: Who received the last three opportunities, and what was the stated reason?

Practice: Write the allocation rule before you allocate. Record who received, who did not, and why. Read that record aloud at review.

Measurable check: A quarterly distribution count of time, money, and opportunity by recipient against stated need. Skew toward insiders is the fail signal (T3).

Regular review: Quarterly, with the people who received least present, not only those who received most.

Where it applies: §4.4, §7.3. Boards, hiring, funding, mentorship slots, and community support.

P12. Moral and spiritual neutrality (§4.7, §6.5)

Looks like: Space for moral reflection and dialogue about values, with your own convictions held personally and lived visibly rather than imposed (§4.7).

Does not look like: Mandatory belief, mandatory practice, or mandatory secular ideology. §6.5 names the failure as worldview dominance, and it infects ideological organizations as easily as religious ones.

Question to ask: Can someone disagree with my deepest conviction and keep their standing in this group?

Practice: Separate the shared floor, meaning conduct everyone has agreed to, from private or tradition-specific commitment. Say which one you are asking for, every time.

Measurable check: At least one member holding a different worldview is present at every decision meeting and has spoken in the last quarter.

Regular review: Every values question is checked against §6.5 at the quarterly review, and any de facto adherence test is removed.

Where it applies: §4.7, §6.5, §9.5. Plural communities, civic organizations, and workplaces.

P13. Bidirectional responsibility with structural guardrails (§4.8, §9.6)

Looks like: Both sides hold duties, and the relationship has boundaries, oversight, and a complaint route that does not pass through the person complained about (§9.6). Effective mentoring is two-way, not a transfer of wisdom from senior to junior (§4.8).

Does not look like: One-way instruction. A grievance process whose first step is to raise it with the subject of the grievance.

Question to ask: If they accused me tomorrow, who would hear it, and what would happen to them next?

Practice: Agree the terms in writing in both directions: what you owe, what they owe, and what either may ask for. Name an alternative recipient for complaints about you and put it where everyone can see it.

Measurable check: Every participant can name the bypass route unprompted, it is used at least once a term, and no complainant pays for using it (§9.6).

Regular review: Annually, the structure is tested by a mock complaint and by asking participants to name the route without help.

Where it applies: §4.8, §9.6. Mentoring programs, coaching, apprenticeships, and workplaces.

Part 2. The five development stages

The stage names are fixed by the Stage 4 brief: beginning, foundation, practice, stewardship, servant leadership. Each ends in a gate, and a failed gate blocks the next stage, following the fail-closed rule the contract locks for stages and tests (architecture §6). These stages are the guide's own arrangement of report material. The guide's ordering follows the locked build sequence S1 to S6 in architecture §6.

Stage 1. Beginning

Entry condition: One named, consensual relationship with a written remit (P1). No prior credential is required, and none may be claimed (R1).

Work: Learn the definition and the failure list until both are recallable (§6, §11). Start P1, P3, and P5 immediately: state the remit, make the monthly ask, use the three-part guidance form. Keep a written log of decisions, refusals, and reversals from the first week. Read §10 and note where the report's own sources are contested, so the weighting is carried forward rather than rediscovered later (C7).

Gate: One full quarter with a remit in force, a log kept, and at least one declined answer recorded (P2).

What blocks advancement: No log. A claimant who cannot show reversals cannot show accountability, and §6.4 treats the absence of review as forfeiture rather than style.

Stage 2. Foundation

Entry condition: Beginning gate passed.

Work: Install the habits that survive motivation. Weekly admission of what you do not know (§6.8). Quarterly referral review (P2). Allocation rules written before allocation (P11). Public correction practised rather than described (§4.2). External feedback requested and treated as non-negotiable, per the report's warning that bias distorts guidance (§9.2).

Gate: Two consecutive quarters where all six tests (Part 3) pass on a real case, judged together with at least one person who is not you (T4).

What blocks advancement: Any §6 pattern visible in the log. Paternalism, dependency-building, or narrative control returns you to the earliest stage the failure touches (§6 preamble).

Stage 3. Practice

Entry condition: Foundation gate passed with a reviewer's confirmation.

Work: Widen the circle only after the guardrails exist. Install the complaint route that bypasses you (P13), a redundancy partner (L4), and written observation limits (P10, §9.1). Put the review with affected people on a fixed calendar (§4.8, §9.2). Begin teaching the three-part form to someone else, since redundancy in leadership is a requirement rather than a courtesy (§9.4).

Gate: A peer runs a guidance conversation in your absence, and the guided person confirms that nothing degraded.

What blocks advancement: Any structure in which you are the only trained person or the only complaint recipient (T3, T4).

Stage 4. Stewardship

Entry condition: Practice gate passed.

Work: Manage what has been entrusted to you, meaning resources, time, and positions, for the benefit of others and the common good, with transparency about flows, regular reporting, and acceptance of oversight (§8, stewardship definition).

Install institutional counters to the elevation problem described in §5.4 and §7.3: distributed decision rights, horizontal accountability channels, and multiple trained leaders rather than one. Publish distribution counts (P11).

Gate: An independent review of resource flows finds no self-benefit pattern, and the review is signed by the people affected and not only by leadership (§6.6, §9.3, L2).

What blocks advancement: Financial, social, or emotional benefit that depends on other people staying dependent (§6.6, §6.7).

Stage 5. Servant leadership

Entry condition: Stewardship gate passed.

Work: Service before authority, and community built through stewardship. That is the Greenleaf content the report uses, carried with the report's own caution that it is a management theory with weak empirical grounding rather than settled science (§7.2, §10, C7). The practical work is obsolescence: handing over decisions, strengthening peer relationships, and reducing your necessity (P8, §3.4).

Gate: The group functions well and decides without you, which is the report's own test that good shepherds make themselves increasingly unnecessary (§3.4, §5.3).

What blocks advancement: Nothing, and that is the point. Standing is forfeitable performance at every stage including this one (§6 preamble, R3). A disqualifying pattern discovered later returns the claimant to the earliest stage the failure touches, and the affected people decide whether the relationship continues (T2, §5.2).

Part 3. The six standing tests

The six tests are locked interpretation from architecture §4, built only from cited report material. Run T1 before acting, T2 before entering, T3 through T6 before scaling, and all six at every scheduled review. The contract's fail-closed rule applies throughout: an unclear case counts as a fail (architecture §6, S3).

T1 Agency test (§3.1, §6.1, §8). After this practice runs, does the choice remain with the person it concerns? Pass when options and risks were shared, the decision stayed with them, and no path was rewarded or punished into preselection. Fail when a decision was taken for their own good.

T2 Consent test (§5.2, §8). Did the relationship enter by agreement, and can it be exited without catastrophic cost? Pass on willing entry, a stated remit, and a clean exit. Fail on imposed shepherdhood or priced refusal.

T3 Power test (§9.1, §6.7, §4.6, §5.4, §7.3). Does the arrangement concentrate dependency, surveillance, or hub status in the shepherd? Pass when information flows openly, outside relationships are welcomed, authority is distributed, and the institutional environment is audited for elevation effects. Fail on deliberate reliance-building or continuous monitoring.

T4 Accountability test (§2.2, §6.4, §9.3, §9.6). Are the shepherd's decisions reviewable by someone other than the shepherd? Pass when review, correction, and bypass channels exist and are actually used. Fail on silencing dissent, controlling narratives, or shielding yourself.

T5 Pluralism test (§6.5, §4.7, §9.5). Is any worldview being imposed, and does disagreement survive? Pass when conviction is held personally, exploration is invited, and dissent is tolerated within the shared values. Fail on demanded conformity, religious or secular alike.

T6 Vulnerability test (§2.2, §9.2, §9.4, §6.8, §5.1, §7.2). Are the shepherd's own fallibility, bias, and limits actively managed, and is care oriented to the weakest member first? Pass on regular self-audit with external feedback, stated confidence levels, redundancy in leadership, and weakest-first orientation. Fail on projected certainty, unaudited bias, or single-point-of-failure structure.

Part 4. The concrete practice set

The Stage 4 brief names eleven practices. Each one below is a procedure, not a virtue, and each carries its citation.

1. Listening

Ask before you tell. Restate the other person's position in their own terms until they confirm it, and do not proceed while the restatement is wrong. Treat silence as untested, never as consent, and ask directly (§3.1, §4.5). Timebox it: in any conversation where you talk more than half the time, stop and hand the floor over (P5).

2. Observation

Record what people do, with a date, and keep the observation separate from your interpretation of it. The report's model is Moses learning pastoral sensitivity by watching one stray goat, which puts careful attention before theory (§4.1, §7.1). Review the log monthly and look for the pattern you kept not writing down.

3. Risk awareness

Name the danger the person cannot yet see, and ask a second observer to check the list, because §9.2 states plainly that bias decides whose risk gets noticed. Put the most vulnerable person first, the order Isaiah's shepherd keeps (§7.2). Separate what you observed from what you assumed.

4. Risk communication

State four things in order: the risk, the evidence, your confidence, and what would change your mind (§4.2, §6.8). Deliver it with regard for the listener's willingness to receive it, which is the difference between frank speech and force (§3.2). Never present a preference as a fact.

5. Risk reduction

Reduce harm without removing the choice. Prefer reversible steps, and leave the cost of refusal acceptable, because anything that makes saying no catastrophic is coercion however gently administered (§8, T2, C2). Agree in advance which signal means you both revisit the plan.

6. Protection without control

Offer protection in a form that can be declined while standing and safety are preserved (§5.2, §8). If declining your protection costs someone their housing, income, membership, or reputation, the design is coercive and you built it (C2, T2). Rebuild the offer until refusal is genuinely affordable.

7. Recognizing limits and uncertainty

Decline where you are unqualified and name someone who is not (§5.1). State confidence levels as numbers or words, not atmosphere (§6.8). Say out loud what would change your position, and mean it (§4.2).

8. Refusing false certainty

Never present speculation as fact. Report the cases where you were wrong to the same audience that heard the original claim (§4.2, §6.8). When someone asks for a guarantee, name the request and decline it, because the promise is the beginning of the failure (§6.8).

9. Refusing manipulation

State your intent before you ask for anything. No love bombing, no guilt induction, no gaslighting, no vulnerability deployed strategically, all of which §6.3 names as manipulation with documented damage. Say what you value openly and let people agree or disagree in the open (§6.3).

10. Refusing dependence

Publish what you know (P9). Welcome outside advisors and never criticize a person for consulting them (§6.7). Hand off relationships and decisions on a schedule, and when your role has served its purpose, say that it is over (§5.3, P8).

11. Review with the people affected

Put it on a calendar rather than leaving it open. Their agenda goes first, decisions and reversals are on the table, and the outcome is recorded (§4.8, §9.2, §9.3, §9.6, L1 through L5). Ask three fixed questions each time: what helped, what harmed, and what should change next quarter.

Part 5. Community and organization

The individual path is the spine, and this part applies it to groups. Institutions carry the same duties and add structure, because §5.4 says institutional arrangements can distort shepherdhood and the shepherd must audit their own environment for that distortion.

Distribute the role, do not appoint it. The report's central hadith applies shepherdhood at every social level precisely so that no one person owns anyone else's fate (§1, §3.1, §7.2). Practically: every function has at least two trained people, and every person in the group has an identified alternative to you (§9.4).

Refuse the elevated single figure. The evidence in §5.4 and §7.3 documents the pattern: physical and organizational elevation creates psychological distance, centralized decisions breed dependency, and isolation blocks the corrective feedback that would have caught the leader's errors.

Structure follows from that finding: decision rights are distributed, feedback channels are horizontal, and no one holds all the answers in one chair.

Price the exit at zero. The line between moral weight and coercion is whether people can leave without catastrophic consequence (§8, T2). Build the exit path before you need it: how to leave, what you keep, what the group owes you, and what you owe the group.

Make complaint routes independent of the complained-of. §9.6 requires channels that bypass the mentor, and §6.4 treats refusal of accountability as forfeiture. Publish the route, rotate who receives complaints, and record the outcome.

Treat pluralism as structure. Communities thrive on pluralism, and one worldview dominating others destroys the diversity that makes collective judgement robust (§9.5). Reserve real seats, with votes, for members who disagree with the founding convictions of the group (P12, §6.5).

Audit concentration of benefit. §4.4 and §7.3 use Ibn Khaldun's finding that concentrated benefits narrow the social solidarity that sustains just governance. Run the distribution count (P11) as an institutional measure, not a private one, and publish the trend.

Run the organization through the same five stages. A group begins with one consensual remit, builds foundation habits, scales only after guardrails exist, then stewards resources under independent review, and finally measures itself by how well it functions without any single leader. An institution that skips to stewardship without P13 guardrails reproduces the failure in §5.4 at scale.

Part 6. The review loop as operating practice

The contract locks these five as operating practice, not as a preface (architecture §7). The calendar above installs them.

L1 (§9.2). Regular self-audit against bias and power effects, with external feedback treated as non-negotiable.

L2 (§9.3). Conflict disclosure and independent review wherever the shepherd benefits from dependence, as the minimum safeguard.

L3 (§9.6). Complaint and correction channels that bypass the shepherd entirely.

L4 (§9.4). Redundancy in leadership through multiple trained people, distributed authority, and peer oversight, because no shepherd is infallible.

L5 (§4.2, §8). Public self-correction as the trust mechanism. Pastoral authority strengthens when proven reliable and weakens when proven wrong, so the record of corrections is the asset.

Part 7. Failure modes and what to do when you find one

The eight disqualifying patterns are locked in architecture §5, mirroring §6.1 through §6.8, and the report's framing is binding: these are structural problems in how a person relates to others, not minor lapses (§6 preamble).

D1 Paternalism (§6.1). Deciding for others for their own good, and rewarding compliance while punishing dissent. The report's antidote is midwifery: help people deliver their own decisions.

D2 Coercion (§6.2, §8). Force, threats, or withheld rewards in place of reason. Apply the cost test: what does refusal actually cost them?

D3 Manipulation (§6.3). Hidden intent, including the four named mechanisms. The alternative is transparent statement of values.

D4 Refusal of accountability (§6.4). Rejecting criticism, silencing dissent, controlling narratives, shielding yourself.

D5 Worldview dominance (§6.5). Imposing a religion, ideology, or philosophy. Converts guidance into indoctrination in either register.

D6 Self-enrichment (§6.6). Financial exploitation, emotional draining of the guided, credit-taking, and resource allocation that favours your own projects.

D7 Dependency engineering (§6.7). Gatekeeping information, discouraging outside relationships, criticizing alternative advisors, framing yourself as uniquely necessary.

D8 False certainty (§6.8). Claiming unpossessed knowledge, presenting speculation as fact, refusing updates against contrary evidence.

The six cross-cutting warnings in §9 attach to all eight as structural amplifiers: surveillance is not care (§9.1), bias distorts guidance (§9.2), conflicts of interest corrupt stewardship (§9.3), fallibility demands redundancy (§9.4), one worldview dominating others destroys community (§9.5), and mentorship power requires guardrails (§9.6).

When you find a pattern in your own record: stop the practice, tell the people affected in the next review rather than in private, hand their decisions back, and let them decide whether the relationship continues (§6, R3, T2).

Do not renegotiate the definition to fit the conduct (§11, C5). A single lapse handled this way is the mechanism the report itself recommends, since standing is re-earned through the visible record and not through assertion (§4.2, §8).

Part 8. What this path still cannot tell you

The report is honest about its own footing, and this guide inherits that honesty rather than smoothing it (§10, C7).

No single tradition or thinker captures the whole of shepherdhood (§10). Greenleaf's servant leadership carries weak empirical grounding and is used here only as a bridge between vocabularies (§10). Foucault's reading of Socrates is contested among classicists (§10).

Xenophon's Cyrus is political romance, not history (§10). The Socrates available to us is a composite built mainly from Plato's early dialogues (§10). The modern mentoring and leadership findings carry the standard limits of social science, including small samples, narrow populations, self-report bias, and correlation standing in for causation (§10).

Two consequences follow for practice. First, interpretations are labelled as interpretations wherever a section of this guide reads as a rule but rests on a contested source, and every passage above that goes beyond literal report wording is marked as such (C6). Second, repetition does not strengthen a weak source. Citing the same tradition five times yields one citation, not five (§10, C7).

Where this guide runs out, it stops. It offers no procedure for a decision that requires a licensed professional, no test for which worldview is true, and no method for being certain about anything (§5.1, C1, C3, §6.8). What remains is the working definition in §11, the thirteen principles, the six tests, and the discipline of asking the people affected whether any of it worked.