Checklists and Forms

Template

Adopt this document under your own document control. Identifiers, roles, and references are placeholders. Nothing here is regulatory advice.

Companion forms to the work instruction and the risk-based assurance strategy. Each form is intended to become a controlled form template in your QMS.

Document identifiers

FRM-CSA-01 … FRM-CSA-06, WI-CSA-01, TQ-001, and SOP-01 are placeholders for your document control system's identifiers. Rename them to match your QMS, and keep the cross-references consistent when you do.

Response convention throughout: Y (satisfied) / N (not satisfied) / N/A (not applicable, with rationale). Any N blocks progression until resolved or a documented exception is approved by QA.

Risk tiers below use the three tiers defined in the risk-based assurance strategy: Tier 1 (high), Tier 2 (medium), Tier 3 (not high). These correspond to the @pytest.mark.gxp_risk("high" | "medium" | "not-high") marker values.

FRM-CSA-01 — Pre-Execution Readiness Checklist

Completed by System Owner; approved by QA before execution.

Field	Entry
System / version
Qualification phase	IQ / OQ / PQ
Protocol tag
Commit SHA (full)
Scheduled execution date

A. Documentation readiness

#	Check	Y/N/NA	Evidence / comment
A1	Validation Plan approved and current, including the risk-based assurance strategy
A2	URS approved as a controlled document
A3	Derived specifications complete for this phase
A4	Every requirement has an explicit expected result
A5	Every requirement carries a `Traces-To:` value resolving to an approved URS requirement
A6	Full commit SHA recorded in the header above, matching `git rev-parse <tag>^{commit}`, and `git tag -v` verifies the tag
A7	FRM-CSA-04 tier classification approved by QA for every requirement in scope
A8	Prior phase closed (OQ not commenced before IQ disposition, etc.) or concurrent execution justified

B. Protocol freeze

#	Check	Y/N/NA	Evidence / comment
B1	All specification and test changes merged to the release branch
B2	Working tree clean; no uncommitted changes
B3	Annotated tag created and pushed
B4	`git describe --exact-match --tags` returns the approved tag
B5	Protocol package assembled (specs, tests, FRM-CSA-04, tag/SHA record)
B6	QA approval of protocol package recorded, dated before execution date

C. Test design

#	Check	Y/N/NA	Evidence / comment
C1	Every requirement in scope has ≥1 linked test via `@pytest.mark.requirements`
C2	Every test's `gxp_risk` marker matches its FRM-CSA-04 tier
C3	Every Tier 1 test captures objective evidence via `gxp_evidence`
C4	Every Tier 1 requirement has ≥1 negative / boundary / error-path case
C5	No test references a requirement ID absent from the specifications
C6	Data integrity requirements (audit trail, access control, immutability) have dedicated tests
C7	Tier 1 executor or reviewer identified and independent of the developer of the function under test

D. Tool readiness

#	Check	Y/N/NA	Evidence / comment
D1	`pytest-gxp` version pinned exactly (`==`) in the lock file
D2	That version qualified under TQ-001 and the qualification is current
D3	Installed version and distribution SHA-256 recorded below
D4	Python version, pytest version, and OS/image digest recorded below
D5	PDF rendering dependencies installed and functional
D6	`--gxp-strict` will be passed at execution, so the high-risk evidence gate is armed

pytest-gxp version : ______________  dist SHA-256: ______________________________
pytest version     : ______________  Python: ____________  OS/image digest: ______

E. Environment

#	Check	Y/N/NA	Evidence / comment
E1	Environment matches that required by the Validation Plan for this phase
E2	For IQ / PQ: execution is against the production deployment
E3	For OQ in staging: environment equivalence rationale approved and evidence attached
E4	Container image digest / package manifest hash recorded
E5	IaC revision recorded
E6	Application, database, and schema versions recorded
E7	Clock synchronised to a documented time source; timezone explicit
E8	Test data prepared; source and any de-identification documented
E9	Access controls in the target environment configured as intended (not relaxed for testing)

F. Credited development testing (complete only if crediting)

#	Check	Y/N/NA	Evidence / comment
F1	SDLC / configuration-management procedure documented and in force
F2	Protected branches technically enforced on the release branch
F3	Code review evidenced per merge by a non-author reviewer
F4	CI logs and coverage output retained for the required period
F5	Test code under the same change control as application code
F6	Specific credited test identifiers named in the traceability matrix (not cited in aggregate)
F7	No Tier 1 requirement is being credited to development testing

Sign-off

Role	Name	Signature	Date
System Owner (prepared)
QA (approved to execute)

FRM-CSA-02 — Post-Execution Report Review Checklist

Completed by the independent reviewer; approved by QA.

Field	Entry
Protocol tag / SHA
Execution date and time (with timezone)
Executed by
Exit code
Report metadata status	☐ FINAL ☐ PROVISIONAL
Report artifact SHA-256 (from `artifact_manifest.sha256`)

A. Execution integrity

#	Check	Y/N/NA	Comment
A1	Execution performed against the approved tag; report source provenance shows the expected commit and tag, and a clean tree
A2	Installed tool version matches the TQ-qualified pinned version, per report metadata `generator`
A3	Environment identifiers in the report match FRM-CSA-01 Section E
A4	Complete console log retained
A5	All execution attempts against this tag are accounted for, including abandoned runs and re-runs to supply deviation references
A6	`artifact_manifest.sha256` verified pre-review and every hash reproduces
A7	Signatory name fields were not populated at generation time
A8	Report timestamps carry an explicit timezone designator
A9	`findings_summary` reviewed; every error and warning finding assessed and dispositioned

B. Result fidelity

#	Check	Y/N/NA	Comment
B1	Total / passed / failed / skipped / errored counts independently reconciled against the test execution register
B2	Reported pass rate arithmetically consistent
B3	Reported coverage and verification rates arithmetically consistent
B4	Counts consistent across CSV, JSON, Markdown, and PDF renderings of the same run
B5	No requirement reported as verified where its verifying test failed, errored, or was skipped
B6	No test that errored is reported as passed
B7	Any result achieved only on repeat execution is recorded, with both outcomes

C. Traceability

#	Check	Y/N/NA	Comment
C1	Every requirement in scope appears in the matrix
C2	No requirement appears in the uncovered list, or each is justified and reclassified
C3	No test cites a requirement ID absent from the specifications (no `unknown-requirement-ref` finding)
C4	Requirement hashes in the matrix match the tagged specification text
C5	Requirement-to-phase mapping correct (`IS-`→IQ, `DS-`/`FS-`→OQ, `US-`→PQ)
C6	Matrix `Risk Tier` values match FRM-CSA-04 for every requirement

D. Evidence

#	Check	Y/N/NA	Comment
D1	Every Tier 1 requirement has ≥1 evidence artifact
D2	Each evidence file referenced in the manifest exists on disk
D3	Each manifest entry's `sha256` matches the stored file
D4	Evidence is attributed to the correct test and requirement
D5	Evidence substantiates the pass determination — reviewer inspected the artifact, not only the result
D6	Evidence timestamps fall within the execution window
D7	Thumbnails are derived images, not substituted for full-size evidence
D8	Evidence contains no unnecessary personal or patient-identifying data

E. Discrepancies

#	Check	Y/N/NA	Comment
E1	Every FAILED result carries a `deviation_ref` in the record, matching the raised deviation
E2	Every SKIPPED / NOT_EXECUTED result has a recorded justification; the preserved skip reason is consistent with it
E3	Every ERROR has a discrepancy record; affected requirements not treated as verified
E4	Retests, where performed, reference the original record and were executed under the approved tag or an approved amendment
E5	No Tier 1 deviation remains open
E6	Report metadata status is `FINAL`; no `PROVISIONAL` banner present

F. Disposition

Disposition	☐ Qualified ☐ Qualified with conditions ☐ Not qualified
Conditions and closure dates
Rationale

Role	Name	Signature	Date
Independent Reviewer
QA (disposition)

FRM-CSA-03 — pytest-gxp Tool Qualification Checklist

Completed on initial qualification and on every version change. Supports TQ-001.

Field	Entry
Currently qualified version
Candidate version
Distribution filename
Distribution SHA-256
Python / pytest versions used for qualification
Reason for re-qualification	☐ initial ☐ version change ☐ dependency change ☐ periodic ☐ defect response

A. Change review

#	Check	Y/N/NA	Comment
A1	Release notes and commit range reviewed between qualified and candidate versions
A2	Changes affecting requirement parsing identified and assessed
A3	Changes affecting result determination identified and assessed
A4	Changes affecting coverage arithmetic identified and assessed
A5	Changes affecting evidence handling identified and assessed
A6	Changes affecting report rendering or metadata identified and assessed
A7	Transitive dependency changes reviewed
A8	Supplier-side records available (repository history, CI results, release process)

B. Automated suite execution

#	Check	Y/N/NA	Comment
B1	Qualification suite obtained from the sdist or tagged release for the candidate version
B2	Candidate installed from a wheel or sdist, not editable, in an isolated environment
B3	`TQ_PINNED_VERSION` and `TZ=UTC` set; suite executed per TQ-001 §6
B4	All mandatory test cases (TQ-1.x through TQ-7.x) passed
B5	Each gap-assessment case yielded its documented expected outcome
B6	Any gap-assessment case that changed state investigated and documented
B7	`tq_console.log`, `tq_evidence.jsonl`, `tq_environment.json`, and `tq_pip_freeze.txt` retained

C. Independence of the qualification

#	Check	Y/N/NA	Comment
C1	Qualification suite execution reviewed by a person other than the author of the pytest-gxp code
C2	Supplier role (package author) and user role (this organisation's qualification) documented as distinct
C3	Where C1 cannot be satisfied, the self-attestation is recorded as a limitation with a stated compensating control
C4	Negative controls confirmed present: failing test surfaces as FAILED; errored test surfaces as ERROR; uncovered requirement surfaces as uncovered; evidence cross-attribution detected

D. Outcome

#	Item	Entry
D1	Disposition	☐ Qualified for use ☐ Not qualified
D2	Accepted limitations carried forward (list each finding and its compensating procedural control)
D3	Procedural controls retired because the corresponding gap is now closed in the tool (list, with the decision reference)
D4	Tool register updated with version and hash
D5	Project lock file pin updated under change control

Role	Name	Signature	Date
Performed by
Reviewed by (independent)
QA approved

FRM-CSA-04 — Requirement Risk Classification Worksheet

One row per requirement. Approved by QA before protocol freeze. Maintain as CSV alongside the specifications, so that classifications can be reconciled against the Risk Tier column of the generated traceability matrix at review.

Column definitions

Column	Content
`requirement_id`	e.g. `FS-014`
`requirement_title`	From the specification heading
`traces_to`	URS requirement ID
`gxp_critical`	Y/N — Q1 per the risk-based assurance strategy
`q1_rationale`	Why failure could / could not compromise safety, quality, or record integrity
`impact`	Direct / Indirect / N/A — Q2 per the risk-based assurance strategy
`compensating_control`	Required if Indirect. Name the specific detection control. "User would notice" is not acceptable.
`detectable_before_reliance`	Y/N — would failure be detected before the record is relied on or product released?
`risk_tier`	1 / 2 / 3
`gxp_risk_marker`	`high` / `medium` / `not-high` — the marker value that must appear on the verifying test
`assurance_method`	Scripted+evidence / Scripted automated / Unscripted / Credited dev testing
`credited_test_ids`	Required if `assurance_method` = Credited dev testing
`negative_testing_required`	Y/N
`downgrade_rationale`	Required if the tier was reduced from the default assignment
`classified_by` / `date`
`qa_approved_by` / `date`

Worked examples

requirement_id	gxp_critical	impact	compensating_control	risk_tier	gxp_risk_marker	assurance_method
`FS-014` Released batch record fields immutable	Y	Direct	—	1	`high`	Scripted + evidence
`FS-021` Audit trail captures user, timestamp, old and new value	Y	Direct	—	1	`high`	Scripted + evidence
`FS-032` Batch report totals recalculated from source on render	Y	Indirect	Independent recalculation by QA reviewer against source data before release, per `<SOP>`	2	`medium`	Scripted automated
`FS-047` Result table sortable by column	N	N/A	—	3	`not-high`	Unscripted
`IS-003` Application deployed at pinned version with expected file inventory	Y	Direct	—	1	`high`	Scripted + evidence

Classification integrity rules

Where the tier is contested, assign the higher tier.
A downgrade from Tier 1 to Tier 2 requires a named compensating control and documented QA concurrence.
detectable_before_reliance = N forces Tier 1 regardless of impact.
Reclassification after execution has begun requires a protocol amendment.
The gxp_risk_marker column and the marker on the verifying test are reconciled at FRM-CSA-02 C6 against the matrix Risk Tier column. A worksheet tier that does not reach the record is not an enforced tier.

FRM-CSA-05 — Unscripted Test Session Record

For Tier 3 assurance, and for supplementary exploratory testing at any tier. One record per session. The session charter is approved before the session begins.

Recording the session as evidence

The fields below can be captured as an evidence artifact inside the run rather than transcribed by hand afterwards. The gxp_evidence fixture provides:

@pytest.mark.gxp
@pytest.mark.gxp_risk("not-high")
@pytest.mark.requirements(["US-012"])
def test_result_table_exploratory_session(gxp_evidence):
    gxp_evidence.record_unscripted_session(
        charter="Explore result table sorting and filtering; a problem is "
                "any ordering that misrepresents the underlying data.",
        tester="A. Tester, QA Analyst",
        duration_minutes=45,
        observations=[
            "Sorted by batch number ascending and descending; order matched source query.",
            "Filtered to a single site; row count reconciled against the database.",
            "Sorted an empty result set; no error, empty table rendered.",
        ],
        defects=["DEF-2026-031: column header tooltip truncated at narrow widths"],
    )

The session record is written as an evidence entry in the manifest with UTC timestamps, attributed to the test that recorded it, and appears in the human-readable report as a labelled list. Use it so the session record is part of the same record set — and the same hash manifest — as the rest of the run. Complete the form below for the fields the API does not carry (charter approval, sign-off, tier reclassification assessment), and cross- reference the evidence entry ID.

Field	Entry
Session ID
Evidence entry ID (from the manifest)
System / version / environment
Requirements or functional area in scope
Risk tier of items in scope
Charter — what is being explored, and what would constitute a problem
Charter approved by / date
Tester (name, role)
Start / end time (with timezone)
Duration (active testing)

Session log — record what was actually done, not what was intended. These rows correspond to the observations list.

#	Area explored	Action taken	Observation	Assessment
1				OK / Anomaly / Defect
2

Findings — these rows correspond to the defects list.

#	Description	Severity	Defect / discrepancy ref	Requirement affected	Tier reclassification needed?

Session conclusion

Field	Entry
Coverage achieved against charter	Full / Partial (state gaps)
Areas deliberately not explored, and why
Any finding indicating a requirement is under-classified
Evidence attached (screenshots, notes, exports)
Conclusion for the requirements in scope	Acceptable / Not acceptable / Further testing required

Role	Name	Signature	Date
Tester
Reviewed by

A session record showing only "explored the module, no issues found" does not constitute assurance evidence. The log must show what was actually exercised.

FRM-CSA-06 — Change Risk Assessment

Completed for every change to the system after initial qualification.

Field	Entry
Change ID / change control ref
Description of change
Requestor / date
Category	☐ application code ☐ configuration ☐ infrastructure / dependency ☐ specification ☐ `pytest-gxp` version ☐ data model

A. Impact assessment

#	Question	Y/N	Detail
A1	Does the change introduce a new requirement?
A2	Does it alter the behaviour verified by any existing requirement?
A3	Does it affect any Tier 1 requirement?
A4	Does it affect audit trail, access control, electronic signature, or record retention behaviour?
A5	Does it affect a calculation, algorithm, or decision rule relied upon for a GxP outcome?
A6	Does it alter the runtime, container base image, or any pinned dependency?
A7	Does it invalidate the OQ environment equivalence rationale?
A8	Does any existing tier classification require revision?
A9	Requirements affected (list IDs and tiers)

B. Routing determination

Condition	Route	Selected
A3 = Y, or A1 = Y introducing a Tier 1 requirement, or A4 = Y, or A5 = Y	Protocol amendment; re-execute affected Tier 1 tests plus full regression; partial revalidation report	☐
A2 = Y affecting Tier 2 only	Full automated regression pass plus this assessment, under change control; no revalidation report	☐
Tier 3 only, or no requirement impact	Standard change control; retain regression pass as evidence	☐
A6 = Y	Re-execute IQ tests plus full regression; reassess environment equivalence	☐
`pytest-gxp` version change	Re-qualify per TQ-001 and FRM-CSA-03 before the next qualification run	☐

A regression pass alone is never sufficient where the change introduces or alters a Tier 1 requirement.

C. Execution and closure

#	Item	Entry
C1	Regression run tag / SHA
C2	Regression outcome (counts, any non-passing result and its record)
C3	Tier 1 tests re-executed (IDs)
C4	Specifications and FRM-CSA-04 updated
C5	Validated state maintained?	Y / N

Role	Name	Signature	Date
Assessed by
QA approved