The Supervisor's Field Kit
What this is: The three working documents that turn "supervision happens" into supervision that can be evidenced: a session record with the standing agenda, a documentation-review rubric scored per record, a monthly case-sampling schedule, and an action log that carries between sessions.
It assumes: Case-carrying staff supervised on a regular cadence, with documented casework.
Change before use: The cadence, sample sizes, and rubric criteria marked [adjust] — each must match your licensing standards and contracts.
This document explains what each of the three documents is for and how to run it. The documents themselves are the accompanying spreadsheet, which has the session record with the standing agenda already laid out, the documentation review rubric as a row per record with the six criteria as scored columns, the monthly sampling schedule, and an action log that carries between sessions.
Read this once. Work in the spreadsheet.
What this kit is for
Nearly every human-services organization can say supervision happens. Far fewer can show what it consists of, that it reaches every worker on a schedule, or that anyone systematically looks at the documentation the organization will one day be judged by. The gap shows up in the same three places: a licensing analyst asks for supervision records and receives calendar entries; a case record surfaces in a complaint or a court proceeding and nobody had read it since it was written; a worker struggles for months before the pattern becomes visible, because each conversation started from scratch.
The three documents in this kit close that gap without adding a system. Each is a page or two, used as-is on paper or pasted into whatever you already use. Together they produce a record: what supervision covered, what the documentation review found, and proof that the looking was systematic rather than occasional.
It is not a performance-evaluation instrument, a clinical supervision model, or a substitute for the supervision standards in your license or contracts. It is the working paper layer underneath those things — the part that makes them evidenceable. Check the cadence and sample sizes against your own requirements before adopting; where a licensing standard is stricter than a default here, the standard wins.
Part one — the standing supervision agenda
One page per worker per session. The power of a standing agenda is not the topics — it is that the same topics recur, so drift becomes visible across sessions and nothing depends on either person’s memory. Sections 1, 5, and 6 are the ones most often missing from supervision that otherwise works.
| Section | What it covers | Prompted by |
|---|---|---|
| 1 · Follow-through | Every action item from the last session, read back one by one: done, in motion, or renegotiated with a new date. Nothing silently disappears. | The previous session’s sheet — which is why the sheet files. |
| 2 · The caseload, by exception | Not every case — the exceptions: new assignments, approaching deadlines, cases in crisis, cases with no contact in [30 · adjust] days, cases the worker is avoiding. The worker nominates; the supervisor adds from the sampling schedule (part three). | The worker’s current case list, brought to the session. |
| 3 · Decisions needed | Anything the worker cannot decide alone: approvals, exceptions, escalations, resource requests. Each leaves the session decided, or with a named decider and a date. | The worker. |
| 4 · Workload and coverage | Caseload count against the ceiling; hours reality against the plan; upcoming leave and who covers. Once a quarter, run the numbers rather than impressions. | Caseload figures — a caseload stress-test sheet if you use one. |
| 5 · Wellbeing, asked directly | Not "how are you" in passing. One direct question in a settled moment: what is the hardest thing you are carrying right now? Secondary trauma accumulates quietly, and the standing agenda is what keeps the question from requiring an occasion. | The supervisor, every session, no exceptions. |
| 6 · Development | One skill or experience being built this quarter, and the concrete step since last session. Training-hour progress against the annual requirement, so December is not a surprise. | The development note started at the worker’s last review. |
| 7 · Actions recorded | Who does what by when — both directions, the supervisor’s commitments included. Signed or initialed by both, filed in the supervision file. | The session itself. |
The record it produces
Date, names, the exception cases discussed by identifier, decisions made, and the action list, on one sheet. Filed per worker, the sheets are the supervision record a reviewer asks for — and section 1 makes each sheet enforce the one before it. A missed session leaves a gap in a numbered sequence, which is exactly the honesty the record needs: reschedule it, do not backfill it.
Part two — the documentation-review rubric
The rubric makes case-record review consistent across supervisors and workers, and it separates two questions that get blurred in casual review: was the work done, and does the record show it? A reviewer sampling your files years from now has only the record. The rubric scores each sampled case entry on six criteria, each rated met / partial / not met, with a note for anything below met.
| Criterion | Met looks like |
|---|---|
| Timeliness | The entry was made within [X business days · adjust to your standard] of the contact or event it records. The entry date and the event date are both visible, so lateness is self-declaring rather than hidden. |
| Completeness | Every required element for this entry type is present — for a visit: date, location, who was present, what was observed, safety or wellbeing assessment where required, and next steps. Your license or contract defines the list; the rubric holds it steady. |
| Accuracy and specificity | Concrete and observable, not conclusory. "The home was appropriate" is a conclusion; what was seen, heard, and checked is a record. Names, dates, and figures match the rest of the file. |
| Required actions evidenced | Anything the standard requires for this case type at this point — the monthly contact, the assessment due, the plan review — is either present or has a documented reason and a new date. Absence with silence is the finding a reviewer writes up. |
| Follow-through | The next steps recorded in the previous entry are addressed in this one: done, attempted, or renegotiated. A file of entries that never reference each other is activity, not casework. |
| Professional tone | Written knowing the client, a judge, or an auditor may one day read it. Respectful language, facts distinguished from impressions, nothing that would embarrass the organization read aloud. |
Using the scores
A single partial is coaching material for section 2 of the next supervision session. A pattern — the same criterion partial across workers, or one worker drifting across criteria — is management information: a training need, a caseload problem, or a form that makes good documentation hard. Score patterns by criterion once a quarter and give the summary a place in your quality-improvement cycle. The point of the rubric is never the grade; it is that the organization reads its own records before someone else does.
Part three — the monthly sampling schedule
Reviewing everything is impossible and reviewing nothing is indefensible, so the defensible position is a sample that is systematic, documented, and biased deliberately toward risk. The schedule below is a starting point — set the numbers to your own volume and requirements, then hold them steady, because a sample size that changes when things get busy is not a system.
| Each month, per supervisor | Default | Why this slice |
|---|---|---|
| Random cases per worker | [2 · adjust] | Drawn by anything arbitrary — every nth case on the alphabetical list, rotating. Randomness is what makes the sample mean something: workers cannot polish for it, and reviewers recognize it as a control. |
| New cases opened this month | [1 per worker] | The first weeks set the pattern for the file, and early omissions compound quietly. |
| High-risk or high-acuity cases | [1–2, standing] | Cases flagged by your own criteria — recent incident, court involvement, new placement, complaint. These files are the likeliest to be read by someone else one day. |
| Dormant files | [1] | No entry in [30+ · adjust] days. Either the work stopped or the recording did; both need to surface. |
| Follow-up from last month | All below "met" | Every entry scored partial or not met last month is re-checked. This is the loop that makes the rubric change practice instead of just describing it. |
The log
One line per sampled case: month, worker, case identifier, entry type, rubric result, and the follow-up owed. Twelve months of that log is a document very few organizations can produce and every reviewer respects — proof that the organization inspects its own work on a schedule, finds things, and closes them. It is also, quietly, the fairest thing in this kit for workers: a systematic sample judges the system, where an occasional spot-check judges whoever happened to be caught by it.
Putting the kit to work
Adopt it in this order: the sampling schedule first, because it feeds the other two; the rubric second, applied to the first month’s sample; the agenda third, with the sample results arriving into section 2. Within one quarter the three documents begin referencing each other, and that interlock — sample feeds rubric, rubric feeds supervision, supervision records follow-through — is what turns three pieces of paper into a supervision system. Set the sample sizes and timeliness standards against your license and contracts before first use, and date the version you adopt.
Free to adapt and use. No registration, no fee, no attribution required. This is a working template, not legal, tax, or accounting advice — the requirements that apply to your organization are confirmed by your own counsel, auditor, and licensing authority.