Back to Details

dora-metrics

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

DORA Metrics

DORA指标

Generate DORA metrics reports using Harness Software Engineering Insights (SEI) via MCP.
通过MCP使用Harness软件工程洞察（SEI）生成DORA指标报告。

Instructions

操作说明

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).
所有DORA指标由单一资源类型提供：
sei_dora_metric
。传入
metric
参数选择具体指标类型：
  • deployment_frequency
  • deployment_frequency_drilldown
  • lead_time
  • change_failure_rate
  • change_failure_rate_drilldown
  • mttr
每次调用DORA指标时的必填参数：
team_ref_id
、
date_start
、
date_end
、
granularity
（DAILY | WEEKLY | MONTHLY）。

Step 1: Get a DORA Metric

步骤1：获取DORA指标

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"
部署频率：
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"
变更前置时间：
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"
变更失败率：
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"
平均恢复时间（MTTR）：
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

步骤2：获取细分数据

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"
部署频率的单部署详情：
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"
变更失败率的单失败详情：
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

步骤3：获取团队数据

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"
列出所有团队：
Call MCP tool: harness_list
Parameters:
  resource_type: "sei_team"
获取团队详情（集成信息、开发人员、集成过滤器）：
Call MCP tool: harness_list
Parameters:
  resource_type: "sei_team_detail"
  team_ref_id: "<team_id>"
  aspect: "developers"   # 或 "integrations" | "integration_filters"

Step 4: AI Metrics (Optional)

步骤4：AI指标（可选）

Call MCP tool: harness_get
Parameters:
  resource_type: "sei_ai_adoption"
Related: 
sei_ai_impact
, 
sei_ai_usage
, 
sei_ai_raw_metric
.
Call MCP tool: harness_get
Parameters:
  resource_type: "sei_ai_adoption"
相关资源：
sei_ai_impact
、
sei_ai_usage
、
sei_ai_raw_metric
。

DORA Benchmarks

DORA基准值

MetricEliteHighMediumLow
Deployment FrequencyMultiple/dayWeekly-MonthlyMonthly-6mo6mo+
Lead Time< 1 hour1 day-1 week1-6 months6mo+
Change Failure Rate< 5%5-10%10-15%> 15%
MTTR< 1 hour< 1 day1 day-1 week1 week+
指标精英级优秀级中等低水平
部署频率每日多次每周-每月每月-半年半年以上
变更前置时间< 1小时1天-1周1-6个月半年以上
变更失败率< 5%5-10%10-15%> 15%
MTTR< 1小时< 1天1天-1周1周以上

Report Format

报告格式

undefined
undefined

DORA Metrics Report

DORA指标报告

Period: <date range> Team: <team or org>
时间范围: <日期范围> 团队: <团队或组织>

Performance Summary

绩效摘要

MetricValueRatingTrend
Deployment FrequencyX/weekHighImproving
Lead TimeX hoursEliteStable
Change Failure RateX%MediumNeeds attention
MTTRX hoursHighImproving
指标数值评级趋势
部署频率X/周优秀级上升
变更前置时间X小时精英级稳定
变更失败率X%中等需要关注
MTTRX小时优秀级上升

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

整体评级: <精英级/优秀级/中等/低水平>

Recommendations

建议

  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
undefined
  1. 变更失败率达X% - 投入测试自动化与代码审核
  2. 前置时间呈上升趋势 - 排查PR审核瓶颈
  3. 考虑使用功能标志解耦部署与发布流程
undefined

SEI Resource Types

SEI资源类型

Resource TypeOperationsDescription
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, getTeam definitions
sei_team_detail
list (+ 
aspect
param: developers / integrations / integration_filters)		Per-team sub-resources
sei_metric
list, getGeneric metrics
sei_productivity_metric
getProductivity metrics
sei_org_tree
list, getOrganization structure
sei_org_tree_detail
list, getOrg tree detail
sei_business_alignment
getBusiness alignment
sei_ai_adoption
getAI adoption metrics
sei_ai_impact
getAI impact metrics
sei_ai_usage
getAI usage metrics
sei_ai_raw_metric
getRaw AI metrics
资源类型操作描述
sei_dora_metric
get（需传入
metric
参数）		包含6种DORA指标类型：deployment_frequency、deployment_frequency_drilldown、lead_time、change_failure_rate、change_failure_rate_drilldown、mttr
sei_team
list、get团队定义信息
sei_team_detail
list（需传入
aspect
参数：developers / integrations / integration_filters）		团队子资源详情
sei_metric
list、get通用指标
sei_productivity_metric
get生产力指标
sei_org_tree
list、get组织架构
sei_org_tree_detail
list、get组织架构详情
sei_business_alignment
get业务对齐指标
sei_ai_adoption
getAI adoption指标
sei_ai_impact
getAI impact指标
sei_ai_usage
getAI usage指标
sei_ai_raw_metric
get原始AI指标

Examples

示例

  • "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
  • "我们的DORA指标表现如何？" - 调用
    sei_dora_metric
    四次，分别传入四个核心
    metric
    参数
  • "跨团队对比DORA指标" - 先列出
    sei_team
    ，再针对每个
    team_ref_id
    调用
    sei_dora_metric
  • "我们的部署频率趋势如何？" - 先调用
    sei_dora_metric
    并传入
    metric: deployment_frequency
    ，再获取细分数据
  • "展示AI adoption指标" - 调用
    sei_ai_adoption
    及相关AI资源

Performance Notes

性能注意事项

  • 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.
  • 必须传入
    team_ref_id
    、
    date_start
    、
    date_end
    、
    granularity
    这些必填参数
  • 生成报告前需收集请求时间范围内的完整指标数据，部分数据会导致结果偏差
  • 需对比多个时间段的指标以识别趋势，而非仅依赖快照数据

Troubleshooting

故障排查

No Metric Data

无指标数据

  • 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
  • 验证SEI集成已配置（Git、CI/CD、问题追踪工具）
  • 确认
    team_ref_id
    属于活跃的SEI团队（可调用
    harness_list resource_type: sei_team
    检查）
  • 检查日期范围是否覆盖集成工具已采集的数据
  • 新增集成后需等待一段时间以完成数据收集与计算

Metrics Seem Incorrect

指标数据异常

  • Verify deployment detection rules in SEI settings
  • Check failure classification criteria
  • Review team member mappings via 
    sei_team_detail aspect: developers
  • 验证SEI设置中的部署检测规则
  • 检查失败分类标准
  • 通过
    sei_team_detail aspect: developers
    查看团队成员映射关系