Chapter 36 — Field or community practice
Lesson map
Lesson 36 published. Map, example, EX106.
flowchart TD
N0["Learning outcomes"]
N1["36.1 Placement"]
N2["36.2 Sanitise"]
N3["36.3 Translation"]
N4["36.4 Conflicts"]
N5["36.5 Journal"]
N6["36.6 Useful output"]
N7["36.7 Leave-behind timer"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6 --> N7
N7 --> EX["Worked example"]
EX --> Q["Exercise EX106"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["36.1 Placement"]
N2["36.2 Sanitise"]
N3["36.3 Translation"]
N4["36.4 Conflicts"]
N5["36.5 Journal"]
N6["36.6 Useful output"]
N7["36.7 Leave-behind timer"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6 --> N7 --> Q["EX106"]
[Learning outcomes]
|
v
[36.1 Placement]
|
v
[36.2 Sanitise]
|
v
[36.3 Translation]
|
v
[36.4 Conflicts]
|
v
[36.5 Journal]
|
v
[36.6 Useful output]
|
v
[36.7 Leave-behind timer]
|
v
[Worked example] --> [Exercise EX106]
Learning outcomes
You will (1) complete a placement-style reflection; (2) sanitise reports; (3) translate academic methods to ops.
36.1 Placement
Shadow TRB assessment operations (synthetic in this course).
Field practice in this course uses a synthetic TRB assessment operations placement so students never touch live customer systems. Shadowing means watching intake, scoping, testing windows and handovers with a notebook of questions. Sam's placement journal forbids inventing war stories that did not happen in the synthetic lane. The goal is operational manners, not secret techniques. Synthetic does not mean casual; the same sanitisation and gift rules still bind.
Each shadow day ends with Sam writing down which decision he watched being made and who had the authority to make it.
36.2 Sanitise
No customer identifiers in academic write-ups.
Academic write-ups must contain no customer identifiers, account numbers, raw IPs or hostnames that embed personal names. Sam runs a checklist and a peer initials it before any upload script accepts the file. A single surname in a hostname once nearly escaped; the peer catch became lunch and a teaching slide. Sanitisation is graded as hard as technical content. Citizen data has no place in academic mirrors even when the story feels educational.
Screenshots are the riskiest artefact: a browser tab title or a terminal prompt can carry a customer name even when the main pane looks clean.
36.3 Translation
Wilson intervals become board-usable sentences.
Placement work includes translating Wilson intervals into sentences a board can fund without losing the uncertainty. Sam practices phrases that keep the range visible instead of collapsing to a heroic percentage. Directors who only hear point estimates approve brittle programmes. Translation drills use the same synthetic results as the comparative project. Statistical caution that cannot be spoken aloud will not survive the boardroom.
A typical drill turns "detected 4 of 7, interval roughly 25 to 84 percent" into "we caught about half, and the true rate could be anywhere from a quarter to most". Directors then ask how many more runs would narrow the range, which is the right question.
36.4 Conflicts
Gift policies and vendor dinners — follow TRB rules.
Community and vendor events follow TRB gift policies without special student exceptions. Sam declines dinners that would look wrong in an after-action review even when they look normal in industry blogs. Conflict rules are part of field practice, not a HR appendix nobody reads. When unsure, he asks the placement supervisor before accepting anything of value. Appearance of influence is treated as seriously as actual influence.
Sam logs any vendor hospitality he declines in the placement journal, so a later review can see the boundary was applied rather than assumed. Small branded items follow the same threshold as meals.
36.5 Journal
Weekly: learned, broken assumption, question.
The weekly journal has three forced lines: what was learned, which assumption broke and what question remains. Victory-only journals teach vanity; broken assumptions teach operations. Sam's supervisor reads the question line and often assigns the next shadow slot from it. Entries stay free of customer identifiers like any other academic artefact. A short honest journal beats a long polished diary that hides confusion.
A typical broken-assumption line reads: "I assumed the SOC saw our start message; it went to an unmonitored alias." That line became a runbook fix the following week, which is exactly what the journal is for.
36.6 Useful output
Leave a runbook improvement behind.
Every placement must leave a concrete runbook improvement the host can keep. Sam's leave-behind was a clarified stop-message template the technicians actually pasted. Tourism with a badge ends when the host file tree is unchanged. Improvements are small on purpose so they can be reviewed and merged. Usefulness is measured by what remains after the student badge stops working.
Sam's change was three lines: the engagement id, the window end time and the SOC alias, placed at the top of the template where tired hands look first. The host merged it after one review and still uses it, which is the evidence that the placement mattered.
36.7 Leave-behind timer
Leave-behind testing uses a timer: the host technician gets fifteen minutes alone. Sam may answer clarifying questions but not touch the keyboard. Keyboard touches mean the leave-behind failed.
The timer is harsh and clarifying. If the technician cannot finish without Sam's hands, the documentation failed even if the tool works. Clarifying answers are allowed; silent takeover is not. Sam watches the clock more than the screen during the test. Hands-off success is the only leave-behind grade that counts.
Failures are written up as documentation defects with the exact step where the technician stalled.
36.8 Sanitisation checklist
Before academic upload, Sam runs a checklist: names, account numbers, IPs, hostnames. A peer initials it.
The checklist is boring on purpose so it gets finished. Peer initials create accountability without a second full review meeting. Sam never initials his own form. Upload automation looks for the initial field before accepting the PDF. Two humans and a script beat one hurried author with good intentions.
The checklist order matters: names first, because they hide in the most places, then account numbers, then network identifiers. Hostnames come last because they need the slowest reading, letter by letter, against the customer list.
36.9 Peer initial box
Sanitisation forms without peer initials are rejected by the academic upload script.
Mechanical rejection removes debates about whether a friend will initial later. Sam learned to book peer time before the upload deadline rather than after. Blank initial boxes are treated like missing signatures on findings. The script's rudeness is a privacy control. If the box is empty, the file simply does not leave the workstation.
The script checks that the initials differ from the author's own and that the peer is on the current cohort list, so a borrowed or self-signed box still fails.
36.10 Surname hostname catch
Peer initials on sanitisation once caught a hostname containing a customer surname. Sam buys the peer lunch. Checklists with teeth make kindness easy.
The catch was small and enormous. A hostname that looked technical still carried a person. Lunch was thanks and a public signal that catches are welcome. Checklists without teeth produce lonely heroes; checklists with teeth produce lunch. Celebrating the catch teaches the next peer to look hard.
The hostname followed an old naming habit where branch servers carried the account manager's surname. After the catch, Sam added a hostname-pattern line to the checklist, and the host team opened a ticket to retire that convention.
36.11 Choose stern peers
Sanitisation peer once was Sam's sternest critic. He chooses stern peers on purpose. Kind peers miss surnames in hostnames.
Kindness and thoroughness are different services. Sam schedules the critic when the file matters most. Friendship reviews still happen for tone; sanitisation wants sharp eyes. Stern peers are thanked specifically so they keep saying yes. Selecting for severity in review is a professional choice, not a social failure.
Sam keeps a short list of who caught what in earlier reviews and books the person whose catches match the artefact's risk: hostnames go to the peer who found the surname, and tables go to the one who spots stray account digits.
36.12 Privacy oath and form gates
Sanitisation peer initials are a required PDF form field; blank fails upload. Sam chooses peers who previously caught his mistakes, not friends who skim. The upload script rejects forms under 10 KB as likely blank. Process pedantry protects citizens whose names must never enter academic mirrors. That sentence is the chapter's privacy oath.
Multiple gates exist because any one gate fails eventually. Prior catchers get priority when Sam picks a peer. Tiny files are suspicious even when fields look filled in a screenshot. The oath sentence is signed again on the vault foyer card. Pedantry here is loyalty to people who will never read the thesis.
36.13 Blank-form size gate
Blank-form size gate once caught a peer who uploaded a photo of a blank table. Pedantry wins.
A photograph of an empty table is an old trick against weak validators. The size gate plus field validation stopped it. Sam documented the attempt so the next validator author would not remove the check as paranoia. Process stories belong in the facilitator guide beside the happy path. Adversarial laziness inside the course is still useful for hardening uploads.
The threshold was set by measuring genuinely completed forms, which never fell under the limit.
36.14 Foyer privacy oath card
Privacy oath sentence is signed on the vault foyer card.
The foyer card is visible to anyone collecting a badge for vault work. Signing is quick and slightly solemn, which helps memory. Sam watches new placements sign before their first shadow day. Cards are archived by term so the ritual has a record. Public signing ties privacy manners to entry, not to an unread handbook chapter page.
Anyone who refuses to sign does not receive vault access that term; in practice the question it raises is a conversation, not a standoff.
Worked example (synthetic TRB)
Sam's sanitised reflection describes a stop drill without naming real customers or staff.
Synthetic case — Riverstone Transit Bank (TRB). The bank, Sam Ito, TrackScope/FindingsDesk/PulseMap and all figures are invented for teaching.
Errors and limits
This book grants no authority to test real banks or payment systems. Labs use documentation-range addresses and synthetic data only; exploit code is never provided.
Exercises
- EX106 — Sanitise. What must academic write-ups remove? Rubric: 3 = customer identifiers / sensitive specifics; 2 = partial; 1 = vague; 0 = none.
- EX107 — Journal trio. What three lines go in the weekly journal? Rubric: 3 = learned, broken assumption, question; 2 = two; 1 = one; 0 = none.
- EX108 — Leave-behind. Name a useful leave-behind. Rubric: 3 = runbook improvement / checklist; 2 = partial; 1 = malware; 0 = none.
Synthesis
Field practice is sanitised reflection plus something the host keeps.
Continue studying — chapter 36 of 40
← Chapter 35 (draft) · Course chapter index · Chapter 37 (draft) →