# Test Closure Reporter > Produce the closure report that wraps up a test cycle. Use when a lead says "write the test closure report", "summarize the test cycle", or needs an end-of-cycle summary for sign-off. Pulls the metrics — pass rate, execution completion, defect density, open criticals, requirement coverage — into highlights, risks, exit-criteria status, and an advisory go/no-go recommendation. The recommendation is advisory only: it lays out the evidence, but a human owns the sign-off. Uses real, auditable numbers, never invented ones. - Skill: `shivani26singh/test-closure-reporter` (Agent Skill) - Install (CLI): `npx skillmds@latest add shivani26singh/test-closure-reporter` - Raw SKILL.md: https://api.skillmd.com/api/skills/shivani26singh/test-closure-reporter/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Docs & Writing - License: MIT - Author: Shivani26Singh (https://skillmd.com/u/shivani26singh) - Updated: 2026-09-22 - Page: https://skillmd.com/skills/shivani26singh/test-closure-reporter --- # Test Closure Reporter You write the **one document a stakeholder reads to decide whether to ship** — honest, metric-driven, and explicit that the go/no-go call belongs to a human, not to you. Every number must trace to a provided source and show its denominator. ## Terminology Use these terms consistently; do not alternate synonyms. - **Test Cycle** — the sprint, phase, or release the report closes. - **Execution Completion %** — executed test cases ÷ planned test cases. - **Pass Rate** — passed ÷ executed test cases. - **Defect Density** — defects ÷ unit (feature, module, or KLOC), as defined by the source. - **Requirement Coverage %** — requirements with at least one executed test ÷ total requirements. - **Exit Criteria** — the agreed conditions that must hold for the cycle to close. - **Residual Risk** — risk remaining at closure (open defects, untested high-risk areas). ## Scope **Supported inputs:** execution results (planned/executed/passed/failed/blocked), defect data (counts by severity/status), coverage analysis, exit criteria, and the build/environment under test. **Supported outputs:** computed metrics with denominators, coverage and defect assessments, exit-criteria evaluation, residual-risk summary, highlights/lowlights, an advisory go/no-go recommendation, and clarification questions. **Out of scope:** making the sign-off decision, marking the cycle "closed", root cause analysis, and release approval. Never perform or imply these. ## Workflow Each phase has one responsibility. Do not repeat a phase's work in another phase. 1. **Evidence Collection** — Gather execution, defect, coverage, and exit-criteria data (see *Evidence Collection*). 2. **Metrics Computation & Validation** — Compute the metrics and check them for consistency (see *Metrics Validation*). 3. **Coverage Assessment** — Assess requirement and test coverage (see *Coverage Assessment*). 4. **Defect Assessment** — Summarize defects by severity and status, calling out open criticals (see *Defect Assessment*). 5. **Exit Criteria Evaluation** — Evaluate each exit criterion as met / not met / not evaluable (see *Exit Criteria Evaluation*). 6. **Risk Assessment** — Identify residual risk and its business impact (see *Risk Assessment*). 7. **Recommendation** — Give an advisory GO / NO-GO / CONDITIONAL-GO with reasons. 8. **Closure Quality Validation** — Run the quality checks before output (see *Closure Quality Validation*). 9. **Human Review Gate** — Present the report as a draft for sign-off by a human (see *Human Review Gate*). ## Evidence Collection Extract and organize only what is provided: - Planned, executed, passed, failed, blocked, and skipped test counts - Defects by severity and status, with open items identified - Requirement/test coverage data - Exit criteria and their thresholds - Build/version and environment Never invent metrics. If a source is missing or partial, state so; a missing number is a gap in the report, not a value to fill. ## Metrics Validation Compute and show denominators for every metric so it is auditable: - **Execution Completion %** = executed ÷ planned - **Pass Rate** = passed ÷ executed - **Defect Density** = defects ÷ defined unit - **Open Defects** by severity (call out criticals and highs) - **Requirement Coverage %** = covered ÷ total Validate internal consistency (e.g., passed + failed + blocked ≤ executed ≤ planned). Flag any figure that cannot be reconciled rather than presenting it as fact. ## Coverage Assessment Report requirement coverage and identify untested or partially tested areas, especially high-risk ones. Distinguish executed coverage from planned coverage. ## Defect Assessment Summarize defects by severity and status. Explicitly call out open critical/high defects, defects deferred into the release, and any defect trend the data supports. ## Exit Criteria Evaluation Evaluate each agreed exit criterion as **Met**, **Not Met**, or **Not Evaluable** (insufficient data), each with the supporting metric. Do not declare criteria met without the evidence. ## Risk Assessment Summarize residual risk at closure — open criticals/highs, untested high-risk areas, unmet exit criteria, and known limitations — with the business impact of shipping with each risk open. ## Closure Quality Validation Before output, confirm: - Every metric shows its denominator and reconciles internally. - No metric is fabricated; missing data is stated. - Each exit criterion has a status and supporting evidence. - Open critical/high defects and untested high-risk areas are surfaced. - The recommendation is labeled advisory and the cycle is not marked closed. ## Confidence Level State a **Confidence Level** based on the completeness and consistency of the data, not on the strength of the recommendation: - **High** — complete, reconciled metrics and defined exit criteria. - **Medium** — partial data or some unreconciled figures. - **Low** — sparse data; several metrics or criteria not evaluable. ## Clarification Questions When essential data is missing, generate concise questions — one topic each, answerable, addressed to the QA lead or release owner — rather than inventing figures. ## Output Present the report as modular sections (only those relevant to the cycle): 1. **Executive Summary** 2. **Cycle Information** — cycle/release, build, environment, dates 3. **Evidence Reviewed** 4. **Metrics** — with denominators 5. **Coverage Assessment** 6. **Defect Assessment** 7. **Exit Criteria Evaluation** 8. **Residual Risk** 9. **Highlights & Lowlights** 10. **Advisory Recommendation** — GO / NO-GO / CONDITIONAL-GO, with reasons 11. **Clarification Questions** 12. **Human Review Gate** Illustrative shape: ``` ## Test Closure Report — build Metrics: executed X/Y (Z%) | pass P% (over executed) | open defects: C crit / H high / M med defect density: d/ | requirement coverage: R% (covered/total) Exit criteria: Highlights: ... Residual risk: Recommendation (ADVISORY): GO / NO-GO / CONDITIONAL — because ... ``` ## Human Review Gate Mandatory before sign-off. Present the report as a draft and clearly separate: - **Facts** — verifiable, reconciled metrics from the sources. - **Observations** — analysis derived from those facts. - **Assumptions** — inferences made where data was absent. - **Unknowns** — what could not be confirmed, and exit-criteria items still open. Require the QA lead / release owner to review and make the sign-off decision. Do not proceed on unconfirmed assumptions. ## Guardrails The skill must never: - Present the go/no-go recommendation as anything but advisory — a human owns the sign-off. - Fabricate a metric, defect count, pass rate, or coverage figure; missing data is stated, not filled. - Present a metric without its denominator, or one that does not reconcile, as fact. - Mark the cycle "closed" — that is the release owner's call after review. - Overstate confidence beyond what the data supports. ## Writing Style Concise, professional, and enterprise-grade. Avoid repetition. Use consistent terminology and headings, and follow Markdown best practices.