Prompt detail

Benchmark Gap Investigation

Investigation of medium or low benchmark metrics to understand which signals deserve contextual attention.

Prompt 01
Analyzed reports
BenchmarksPR Cycle TimeIssue Cycle TimeThroughputDORA MetricsCode ReviewPlanning AccuracyEngineering InvestmentProductivityAging BranchesAging IssuesDeploy
Context to provide

Provide the squad, a review period (90 days works well), and an optional repository. Add any benchmark area you already want to understand; otherwise the analysis will inspect every covered metric before narrowing the investigation.

Copyable prompt
3,333 characters
ROLE Engineering consultant using MCP DevStats. OBJECTIVE Use benchmark tiers to show which squad delivery signals deserve investigation now. Deliver a short diagnostic map with evidence, limits, and specific context questions. Do not create a plan, choose priorities, set targets, or judge people. ...

How the analysis runs

The review starts with external benchmark signals, validates each relevant gap with the matching operational report, and finishes with bounded questions instead of a team score.

  1. 01Scope and inputsEstablish comparable benchmark scopeConfirm Squad-Name, the period, repository filters, units, and benchmark coverage so every tier and supporting metric refers to the same population.View detailsHide details
    Confirm Squad-Name, the period, repository filters, units, and benchmark coverage so every tier and supporting metric refers to the same population.
  2. 02Analysis and evidenceValidate the gaps worth inspectingStart with Benchmarks, retain only Medium or lower signals, and query the smallest matching report needed to verify each displayed symptom.View detailsHide details
    Start with Benchmarks, retain only Medium or lower signals, and query the smallest matching report needed to verify each displayed symptom.
  3. 03Interpretation limitsFrame evidence and context questionsConnect corroborated signals without declaring cause, distinguish snapshots from trends, and turn remaining uncertainty into specific questions for the team.View detailsHide details
    Connect corroborated signals without declaring cause, distinguish snapshots from trends, and turn remaining uncertainty into specific questions for the team.

Expected Output

Markdown · Illustrative example

Benchmark Gap Investigation — Squad-Name — Apr–Jun 2026

Scope and coverage

90 days · all repositories · benchmark source: DevStats peer cohort · 11 of 12 metrics covered.

Benchmark signals to inspect

MetricTierEvidenceWhy inspect
PR Cycle TimeMedium6.8 days; peer median 4.2Review stage represents 46% of elapsed time
Change Failure RateLow18%; peer median 11%5 of 28 deployments were followed by a failure

Diagnostic read

  • Review time increased from 2.1 to 3.1 days while PR size remained stable; this is compatible with queueing and deserves context.
  • Deployment volume stayed near 9 per month, while failed changes rose from 2 to 5; this does not identify a technical cause.

Interpretation limits

  • Benchmark tiers are external reference points, not a team score or target.
  • One repository had incomplete incident linkage for 12 days.

To understand the context

  • Where did review queues concentrate during May and June?
  • What changed in incident classification or release conditions during the period?

Evidence used

  • Benchmarks — identify covered tiers
  • PR Cycle Time — validate review-stage concentration
  • DORA Metrics — validate stability signal

Stop guessing. Start asking.

Connect DevStats MCP to your AI assistant and turn these prompts into instant answers. Copy a prompt and go.

Get a demoStart free trial