Set a p95 latency budget, decompose per-stage latency, and tune relevance within it. Reach for this on a speed/capacity question.
Scanned 9/23/2026
npx -y skills add mcorbett51090/RavenClaude --skill set-latency-budget --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Set Latency Budget?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/mcorbett51090-set-latency-budget)More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.
---
name: set-latency-budget
description: "Set a p95 latency budget, decompose per-stage latency, and tune relevance within it. Reach for this on a speed/capacity question."
---
# Skill: Set latency budget
Don't chase NDCG into a timeout (§3 #4).
## Step 1 — Set the p95 budget
A target the search must hold via `search_relevance_engineering_calc.py latency-budget` (§3 #4).
## Step 2 — Decompose per stage
Query, fetch, rescore — find the slow stage (§3 #4).
## Step 3 — Size the index
Shards/replicas/storage via `search_relevance_engineering_calc.py index-sizing` (§3 #7).
## Step 4 — Tune relevance within budget
Buy relevance with rescoring only inside the headroom (§3 #4 #5).
## Output
A p95 latency budget with per-stage decomposition and the relevance headroom. Traverse Tree 3 in the decision-trees file.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!