dora-metrics

Generate DORA metrics and engineering performance reports using Harness SEI via MCP. Track deployment frequency, lead time, change failure rate, and MTTR. Use when user says "DORA metrics", "deployment frequency", "lead time", "engineering metrics", or asks about team performance.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

DORA Metrics

Name: dora-metrics
Rating: 68 (1 reviews)
Author: harness

Generate DORA metrics reports using Harness Software Engineering Insights (SEI) via MCP.

Instructions

All DORA metrics are served by a single resource type: sei_dora_metric. Pass the metric parameter to select the variant:

deployment_frequency
deployment_frequency_drilldown
lead_time
change_failure_rate
change_failure_rate_drilldown
mttr

Required inputs on every DORA call: team_ref_id, date_start, date_end, granularity (DAILY | WEEKLY | MONTHLY).

Step 1: Get a DORA Metric

Deployment Frequency:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "deployment_frequency"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "WEEKLY"

Lead Time for Changes:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "lead_time"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "WEEKLY"

Change Failure Rate:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "change_failure_rate"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "WEEKLY"

Mean Time to Recovery:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "mttr"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "WEEKLY"

Step 2: Get Drilldown Data

Per-deployment detail for frequency:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "deployment_frequency_drilldown"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "DAILY"

Per-failure detail for CFR:

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_dora_metric"
  metric: "change_failure_rate_drilldown"
  team_ref_id: "<team_id>"
  date_start: "2026-03-01"
  date_end: "2026-04-01"
  granularity: "DAILY"

Step 3: Get Team Data

List teams:

Call MCP tool: harness_list
Parameters:
  resource_type: "sei_team"

Get team details (integrations, developers, integration filters):

Call MCP tool: harness_list
Parameters:
  resource_type: "sei_team_detail"
  team_ref_id: "<team_id>"
  aspect: "developers"   # or "integrations" | "integration_filters"

Step 4: AI Metrics (Optional)

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_ai_adoption"

Related: sei_ai_impact, sei_ai_usage, sei_ai_raw_metric.

DORA Benchmarks

Metric	Elite	High	Medium	Low
Deployment Frequency	Multiple/day	Weekly-Monthly	Monthly-6mo	6mo+
Lead Time	< 1 hour	1 day-1 week	1-6 months	6mo+
Change Failure Rate	< 5%	5-10%	10-15%	> 15%
MTTR	< 1 hour	< 1 day	1 day-1 week	1 week+

Report Format

## DORA Metrics Report

**Period:** <date range>
**Team:** <team or org>

### Performance Summary

| Metric | Value | Rating | Trend |
|--------|-------|--------|-------|
| Deployment Frequency | X/week | High | Improving |
| Lead Time | X hours | Elite | Stable |
| Change Failure Rate | X% | Medium | Needs attention |
| MTTR | X hours | High | Improving |

### Overall Rating: <Elite/High/Medium/Low>

### Recommendations
1. CFR at X% - invest in test automation and code review
2. Lead time trending up - look at PR review bottlenecks
3. Consider feature flags to decouple deploy from release

SEI Resource Types

Resource Type	Operations	Description
`sei_dora_metric`	get (+ `metric` param)	All 6 DORA variants: deployment_frequency, deployment_frequency_drilldown, lead_time, change_failure_rate, change_failure_rate_drilldown, mttr
`sei_team`	list, get	Team definitions
`sei_team_detail`	list (+ `aspect` param: developers / integrations / integration_filters)	Per-team sub-resources
`sei_metric`	list, get	Generic metrics
`sei_productivity_metric`	get	Productivity metrics
`sei_org_tree`	list, get	Organization structure
`sei_org_tree_detail`	list, get	Org tree detail
`sei_business_alignment`	get	Business alignment
`sei_ai_adoption`	get	AI adoption metrics
`sei_ai_impact`	get	AI impact metrics
`sei_ai_usage`	get	AI usage metrics
`sei_ai_raw_metric`	get	Raw AI metrics

Examples

"How are we doing on DORA metrics?" - Call sei_dora_metric four times with each primary metric
"Compare DORA across teams" - List sei_team, then call sei_dora_metric per team_ref_id
"What's our deployment frequency trend?" - Get sei_dora_metric with metric: deployment_frequency, then drilldown
"Show AI adoption metrics" - Get sei_ai_adoption and related AI resources

Performance Notes

Always pass team_ref_id, date_start, date_end, granularity — these are required.
Gather metrics across the full requested time range before generating the report. Partial data skews results.
Compare metrics across multiple time periods to identify trends, not just snapshots.

Troubleshooting

No Metric Data

Verify SEI integrations are configured (Git, CI/CD, issue tracking)
Confirm team_ref_id belongs to an active SEI team (harness_list resource_type: sei_team)
Check the date range covers data the integrations have ingested
Allow time for data collection and calculation after new integrations are added

Metrics Seem Incorrect

Verify deployment detection rules in SEI settings
Check failure classification criteria
Review team member mappings via sei_team_detail aspect: developers

Repository: harness/harness-ai
Path: plugins/claude/skills/dora-metrics/SKILL.md
Commit: 23e9b8b

Last updated: 5 days ago
First committed: 6 months ago

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.