Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

SLO Implementation · dokhacgiakhoa | DecimalAI

dokhacgiakhoa/slo-implementation

NewSkillScoreMaintained

Define and implement Service Level Indicators (SLIs) and Service Level Objectives (SLOs) with error budgets and alerting. Use when establishing reliability targets, implementing SRE practices, or measuring service performance.

DevOpsby dokhacgiakhoa

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

DevOps & SRE pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/dokhacgiakhoa-slo-implementation/SKILL.md

decimalai skills pull dokhacgiakhoa-slo-implementation --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/dokhacgiakhoa-slo-implementation/SKILL.md https://app.decimal.ai/s/dokhacgiakhoa-slo-implementation/SKILL.md

Raw file: /s/dokhacgiakhoa-slo-implementation/SKILL.md ↗ · machine-readable summary at /s/dokhacgiakhoa-slo-implementation.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

25 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 6d ago

No measurable change+2% tokens0% turns12% of whole cases pass with skill (3/25)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-23	✓→✓	＝ Same ✓	21%	0%
case-24	✓→✓	＝ Same ✓	16%	0%
case-25	✓→✓	＝ Same ✓	3%	0%
case-01	✗→✗	＝ Same ✗	15%	0%
case-02	✗→✗	＝ Same ✗	6%	0%

Showing 5 of 25 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

dokhacgiakhoa / agent-skills-4-vibe-coding-cli

.agent/skills/slo-implementation/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/10/2026Version v1

Skill contents

SKILL.md

SLO Implementation

Framework for defining and implementing Service Level Indicators (SLIs), Service Level Objectives (SLOs), and error budgets.

Do not use this skill when

The task is unrelated to slo implementation
You need a different domain or tool outside this scope

Instructions

Clarify goals, constraints, and required inputs.
Apply relevant best practices and validate outcomes.
Provide actionable steps and verification.
If detailed examples are required, open resources/implementation-playbook.md.

Purpose

Implement measurable reliability targets using SLIs, SLOs, and error budgets to balance reliability with innovation velocity.

Use this skill when

Define service reliability targets
Measure user-perceived reliability
Implement error budgets
Create SLO-based alerts
Track reliability goals

SLI/SLO/SLA Hierarchy

SLA (Service Level Agreement)
  ↓ Contract with customers
SLO (Service Level Objective)
  ↓ Internal reliability target
SLI (Service Level Indicator)
  ↓ Actual measurement

Defining SLIs

🧠 Knowledge Modules (Fractal Skills)

1. Common SLI Types

2. Availability SLO Examples

3. Choose Appropriate SLOs

4. Error Budget Formula

5. Error Budget Policy

6. Prometheus Recording Rules

7. SLO Alerting Rules

8. Weekly Review

9. Monthly Review

10. Quarterly Review

Want to see if dokhacgiakhoa/slo-implementation would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/30/2026, 08:43 UTC · 25 cases

Lift

12% vs 12%

Agent turns

1.0 vs 1.0

Tokens

+2%

2,803.84 vs 2,749.4

pass → pass · 3fail → fail · 22

Per-case results (25)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 We are launching a critical customer-facing API service and need to set up formal reliability targets. Can you guide us through defining measurable indicators, setting reasonable objectives, calculating our allowed failure allowance, and creating a step-by-step implementation plan with verification strategies?	fail→fail	23,092	22,651	-2%	1	1	0%	3,425	3,947	+15%	0	0	—
case-02 Our engineering team wants to transition from basic infrastructure alerts to service-level objective monitoring using Prometheus. Please provide actionable implementation steps, including how to structure recording metrics, build alerting rules based on target consumption rates, and establish review cadences for our team.	fail→fail	30,608	23,327	-24%	1	1	0%	4,038	4,281	+6%	0	0	—
case-03 We need to design a operational reliability framework for our e-commerce checkout service. Please provide a complete guide covering the structural hierarchy between customer agreements, internal targets, and actual metrics, along with operational policies for managing our error allowance and validating the outcomes.	fail→fail	39,217	29,434	-25%	1	1	0%	7,285	5,427	-26%	0	0	—
case-04 Our product team is confused about the difference between SLAs, SLOs, and SLIs for our online payment gateway. Can you explain how these three terms relate hierarchically from customer-facing commitments down to raw metrics?	fail→fail	20,397	14,938	-27%	1	1	0%	2,750	2,314	-16%	0	0	—
case-05 Our REST API processed 10,000,000 requests last month with an internal SLO target of 99.9% availability. How many failed requests can we tolerate before our error budget is completely exhausted?	fail→fail	9,147	4,110	-55%	1	1	0%	869	1,248	+44%	0	0	—
case-06 Our microservice team has burned through 100% of its quarterly error budget in the first month. What operational policy actions should be triggered when an error budget is fully consumed?	fail→fail	19,375	18,707	-3%	1	1	0%	2,296	2,601	+13%	0	0	—
case-07 We want to measure availability for our HTTP API microservice using server response codes. How should we define an availability SLI as a mathematical ratio of successful requests to total requests?	fail→fail	22,867	14,043	-39%	1	1	0%	2,982	2,080	-30%	0	0	—
case-08 Our frontend team wants an SLI to measure endpoint response times for our search service. How should a latency SLI be structured to capture user-perceived performance rather than average response times?	fail→fail	14,830	16,690	+13%	1	1	0%	2,486	2,490	+0%	0	0	—
case-09 We are writing Prometheus configuration for a gRPC user service. How do we construct a recording rule to compute a 5-minute rolling HTTP request error rate from raw metric counters?	fail→fail	15,244	16,972	+11%	1	1	0%	2,896	2,857	-1%	0	0	—
case-10 Instead of triggering alerts immediately on a single failed request, we want to build alert rules based on error budget burn rate over multiple evaluation windows. What strategy should we use for alert thresholds?	fail→fail	18,711	18,609	-1%	1	1	0%	2,465	3,077	+25%	0	0	—
case-11 We are establishing a weekly operational meeting for our SRE team to review reliability. What key metrics and burn rate signals should be examined during a weekly review?	fail→fail	16,983	13,860	-18%	1	1	0%	2,740	2,815	+3%	0	0	—
case-12 We need an operational template for our monthly service review. What long-term trends and budget consumption figures should be evaluated at the end of each month?	fail→fail	23,096	19,952	-14%	1	1	0%	2,933	2,959	+1%	0	0	—
case-13 At the end of Q2, our leadership wants to evaluate our existing reliability targets against business strategy. What actions should be taken during a quarterly SLO review?	fail→fail	14,808	15,447	+4%	1	1	0%	2,348	2,045	-13%	0	0	—
case-14 Our team is tempted to set a 99.999% ('five nines') availability target for our internal admin dashboard. How should we determine whether this target is appropriate or counterproductive?	fail→fail	17,096	15,794	-8%	1	1	0%	2,824	3,007	+6%	0	0	—
case-15 We have defined Prometheus alert rules for our service's error budget burn rates. How can we formally verify and test that our Prometheus alerting configuration is syntactically valid and firing correctly?	fail→fail	16,992	19,429	+14%	1	1	0%	3,128	3,182	+2%	0	0	—
case-16 We run a nightly data pipeline that processes batch files. How should we measure the reliability of this batch pipeline using job completion time and freshness SLIs?	fail→fail	23,064	24,173	+5%	1	1	0%	3,260	3,716	+14%	0	0	—
case-17 When setting up burn rate alerts for a 30-day SLO of 99.9%, how do short lookback windows (e.g., 5 minutes) and long lookback windows (e.g., 1 hour) combine to avoid false positives and delayed alerts?	fail→fail	15,894	18,906	+19%	1	1	0%	2,900	3,170	+9%	0	0	—
case-18 Our payment processing service has an SLO stating that 95% of requests must complete in under 200ms over a 30-day window. If we receive 2,000,000 requests in 30 days, what is our latency error budget?	fail→fail	9,389	9,282	-1%	1	1	0%	853	1,305	+53%	0	0	—
case-19 Our order entry service suffered a 10-minute outage, but because our monthly SLO is 99.0% and overall uptime remains at 99.4%, our error budget is not exhausted. Should feature deployments be frozen?	fail→fail	16,147	16,097	-0%	1	1	0%	1,785	2,265	+27%	0	0	—
case-20 We operate an Apache Kafka event streaming cluster. What is an appropriate user-centric SLI to measure consumer processing lag for our event processors?	fail→fail	14,377	17,445	+21%	1	1	0%	2,418	2,546	+5%	0	0	—
case-21 Our customer agreement promises 99.5% availability with financial penalties if breached. What internal SLO threshold should engineering target to ensure we do not breach our external SLA?	fail→fail	18,379	11,014	-40%	1	1	0%	2,332	2,359	+1%	0	0	—
case-22 We just deployed error budget tracking dashboards in Grafana connected to Prometheus. What steps should our SRE team perform to verify that the error budget calculation is working accurately in practice?	fail→fail	22,706	21,014	-7%	1	1	0%	3,089	2,980	-4%	0	0	—
case-23 We want to auto-scale our Kubernetes deployment based on high CPU and memory utilization. Can you write an HPA resource manifest targeting 70% CPU and 80% RAM utilization?	pass→pass	7,353	11,917	+62%	1	1	0%	1,500	1,816	+21%	0	0	—
case-24 Our DevOps team needs to configure a PagerDuty escalation policy so that unacknowledged server alerts escalate from primary on-call to secondary on-call after 15 minutes. Can you show us how to set up this escalation policy in PagerDuty?	pass→pass	16,433	11,220	-32%	1	1	0%	2,289	2,666	+16%	0	0	—
case-25 Our production database suffered a 30-minute outage yesterday due to deadlocks during a migration. Can you provide a post-mortem incident template using the 5 Whys analysis framework to document the root cause?	pass→pass	21,850	19,849	-9%	1	1	0%	2,844	2,943	+3%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted. The headline lift of 0 percentage points is the difference between those two pass rates over the 25 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 12%
Pass rate with the skill: 12%
Cases attempted: 25
Cases counted toward lift: 25
Cases that regressed: 1
Skill version measured: v1
Run completed: 8/30/2026, 08:43 UTC

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+12% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested 5 days ago

Other measured skills in the registry, with their headline benchmark lift.