This skills matrix template compares required task levels with demonstrated work, while keeping self-reports and unassessed skills separate. The filled example identifies a specific development need without ranking employees or assigning work automatically.
Define the task before choosing a level
The original fictional team prepares digital archive handoffs. Its three tasks are S1 · Filename checks, S2 · Caption/source matching, and S3 · Handoff manifest preparation. No actual employees, qualifications, completed training, or personnel assessments are represented.
AIHR's skills-matrix guidance connects required and available skills with explicit assessment methods and proficiency levels. Define a scale relevant to the work you need; a number without a description is difficult to interpret consistently.
The original scale here has these meanings:
| Level | Observable work in this exercise |
|---|---|
| L1 · Supported | Completes the supplied task with step-by-step guidance. |
| L2 · Routine independent | Completes a routine supplied case using the agreed checklist without step-by-step help, and flags exceptions. |
| L3 · Exception handling | Handles a supplied exception using the given rules, explains the decision, and escalates what those rules do not resolve. |
| Unknown | No adequate reviewed demonstration is supplied for this task. It is not a level below L1. |
Apply those anchors to the actual task. For filename checks, an L2 demonstration applies the supplied naming rules to the batch and flags a conflicting name. For caption/source matching, it matches caption references to their source records and flags missing support. For a handoff manifest, it lists the supplied files and versions and reconciles them against the source inventory.
L3 does not grant authority to invent a naming rule, change a source, or approve a handoff. The exercise distinguishes how someone handles supplied work from the permissions their real role carries.
Read the filled matrix with its evidence
| Person | S1 · Filename checks | S2 · Caption/source matching | S3 · Handoff manifest |
|---|---|---|---|
| Ana | L2 · E1 | L2 · E4 | L2 · E6 |
| Bo | L2 · E2 | L1 · E5 | L1 · E7 |
| Cy | L1 · E3 | Unknown | Unknown · self-report E8 kept separately |
The complete fictional evidence set is:
- E1: Ana applied the supplied filename checklist to a routine batch without step-by-step help and flagged a conflicting name.
- E2: Bo did the same filename task without step-by-step help and flagged a conflicting name.
- E3: Cy completed the filename task with a guide walking through the steps.
- E4: Ana matched the supplied caption references to source records without step-by-step help and flagged a missing reference.
- E5: Bo completed that matching task with step-by-step guidance.
- E6: Ana prepared and reconciled the routine handoff manifest without step-by-step help, identifying a missing inventory item.
- E7: Bo prepared the routine manifest with step-by-step guidance.
- E8: Cy reports being able to prepare a manifest independently. No reviewed demonstration accompanies that report.
No demonstration is supplied for Cy's caption/source matching. E8 is useful information for planning a review, but it does not become an observed L2 rating. Select Cy / S3 in the example to see the Unknown current assessment beside the separate self-report.
Compare the matrix with a stated coverage target
For the next cycle, the fictional team proposes at least two people at L2 or above for S1, one for S2, and two for S3. These are supplied planning targets, not universal staffing ratios.
| Task | Proposed coverage target | People with supplied L2-or-higher evidence | Review finding |
|---|---|---|---|
| S1 | 2 | Ana and Bo: 2 | The evidence count meets the supplied target. |
| S2 | 1 | Ana: 1 | The evidence count meets the supplied target. |
| S3 | 2 | Ana: 1 | One further evidence-backed L2-or-higher assessment is needed to meet the target. |
Choose Review S3 coverage to inspect E6, E7, and E8 beside the target. Bo has evidence for supported work. Cy has a self-report awaiting review. Those point to different next steps: supported practice for Bo, and an appropriately arranged demonstration to assess Cy's report.
The matrix does not decide who should take either opportunity. Discuss interest, availability, support, and the review process with the people involved. Keep interest separate from current proficiency. Also check capacity and actual role permissions before using a coverage count in an operational plan.
Use the gap to plan a fair next review
For Bo, agree a practice task that moves from step-by-step guidance toward the supplied routine checklist. For Cy, agree what demonstration would provide useful evidence, without assuming in advance that the outcome will be L2. Give both people the task criteria before the review.
Record the assessment date, task version, reviewer, evidence reference, and method when actual reviews occur. Update a cell only when the maintained evidence supports the change. A course attendance record can help explain preparation, but it cannot substitute for the task demonstration this scale requires.
Avoid adding the level codes into an overall employee score. The cells concern different tasks; their purpose is to explain coverage and development questions. A row total would obscure which task the person can perform under the supplied conditions.
Copy a prompt for the task-specific matrix
Create one complete HTML skills-matrix review page.
Team/project tasks and required coverage: [PASTE]
Observable level anchors for each task: [PASTE]
Permitted current assessments with method, date, reviewer, evidence: [PASTE]
Self-reports, interest and review needs, kept separately: [PASTE]
Actual permissions and personnel-record process: [PASTE]
Use supplied evidence only. Define the tasks and scale before
showing ratings. Unknown is not zero or a failed assessment.
Separate self-report, training attendance, observed demonstration
and role authorization. Do not invent competence or employee ranks.
Include this original fictional archive-team example:
Tasks S1 filename checks; S2 caption/source matching; S3 manifest.
L1 completes supplied task with step-by-step guidance.
L2 completes routine case using agreed checklist without step-by-step
help and flags exceptions. L3 handles a supplied exception with given
rules, explains the decision and escalates unresolved issues.
Unknown: no adequate reviewed demonstration supplied.
S1 L2 anchor: apply naming checklist to batch and flag name conflict.
S2 L2 anchor: match caption/source references and flag missing support.
S3 L2 anchor: list files/versions and reconcile to source inventory.
Ana: S1 L2 E1; S2 L2 E4; S3 L2 E6.
Bo: S1 L2 E2; S2 L1 E5; S3 L1 E7.
Cy: S1 L1 E3; S2 Unknown; S3 Unknown with separate self-report E8.
E1 Ana applies routine filename checklist independently, flags conflict.
E2 Bo does same independently, flags conflict.
E3 Cy completes filename task with step-by-step guide.
E4 Ana matches caption/source records independently, flags missing ref.
E5 Bo matches caption/source records with step-by-step guidance.
E6 Ana prepares/reconciles routine manifest independently, identifies
missing inventory item. E7 Bo prepares manifest with step-by-step help.
E8 Cy self-reports independent manifest skill; no reviewed demonstration.
No Cy S2 demonstration. These are invented teaching records only.
Next-cycle proposed L2-or-higher coverage: S1 two people; S2 one; S3 two.
Worked counts: S1 Ana/Bo=2; S2 Ana=1; S3 Ana=1, one short of target 2.
Counting evidence does not establish availability, permission or staffing.
Cy/S3 selection shows Unknown separately from E8 self-report.
Review S3 coverage shows target 2, current 1, E6/E7/E8 and the distinction
between Bo's supported practice need and Cy's demonstration review.
Do not auto-upgrade a level, select a trainee or assign operational work.
Provide matrix, task anchors, full evidence list and the two selected
review views. Include a blank matrix/evidence scaffold and temporary
review questions; preserve typing across views and confirm reset.
No employee score, ranking, course completion as skill or certification.
At least 16px text, native labeled controls, visible focus, stacked
person/task records at 360px without overflow. Inline CSS/JavaScript,
no external assets, network or storage. State notes are temporary.
Return complete HTML; actual personnel records stay in approved systems.
Check the generated S3 finding by hand: only Ana has a supplied L2-or-higher demonstration, against a target of two. The page should keep Cy's self-report visible without counting it as another demonstrated L2 skill.
Review the task anchors before changing a rating
A sanitized HTML matrix can put a task definition, evidence references, and coverage question together. Put it in your NOS workspace when a team lead and reviewer need to agree what a routine-independent manifest demonstration should include.
If “Independent” is too vague, clarify the static scale description in visual editing mode: “Completes a routine case with the agreed checklist, without step-by-step help, and flags exceptions.” Save the wording so colleagues can reopen the existing link and read it. Update the maintained scale and assessment process too; an explanatory text edit does not reassess anyone or authorize work.
Agree the next evidence-gathering or development step with the people involved. Revisit the coverage question after the actual review, retaining the task version and evidence behind each updated cell.