simulate-before-launch

An offline gate for checking a proposed AI model or routing change against a fixed set of test tasks before it reaches production.

In plain words
What is it for?
Use it to replay frozen workloads, score quality and output formats over repeated runs, and produce a ship-or-no-ship decision for one proposed change.
Why use it?
It can reveal wrong answers or incorrectly formatted responses that normal uptime and latency monitoring may miss.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/understudylabs/understudy-agent-tools/simulate-before-launch
Any agent
npx skills add understudylabs/understudy-agent-tools --skill simulate-before-launch
Clone the repo
git clone --depth 1 https://github.com/understudylabs/understudy-agent-tools

Made for: Claude Code, Codex.

Per session 78 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,083 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00078 $0.02083
Opus 5 $0.00039 $0.01042
Sonnet 5 $0.00016 $0.00417
Haiku 4.5 $0.00008 $0.00208

Measured 2d ago against content hash 35716e299806, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

simulate-before-launch scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/simulate-before-launch/SKILL.md · 151 lines

How it starts

The opening of the file, as written. The whole thing — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Simulate Before Launch

Live monitoring watches latency and HTTP errors; a contract regression returns HTTP 200 with the wrong shape. A routed model that answers a response_format: json_schema request with a fenced ```json code block instead of an enforced structured object shows zero errors and great latency while one in five downstream parses fails. This worker is the offline gate between "a change exists" and "traffic moves": replay the workload's frozen tasks through the same serving path with the change applied, score quality and output contracts over repeated rollouts, and emit a launch verdict — before any traffic, capture, or customer sees the change.

Checked against existing skills: design-simulated-environment owns building the seeded environment and validator — this skill runs an existing eval surface as a ship/no-ship gate. capture-evidence owns the frozen harness/metric/splits/baseline this gate refuses to run without. compare-model-sweep owns choosing among many candidates — this skill judges one proposed change against the incumbent. ramp-and-verify owns live traffic and consumes this skill's verdict as its "frozen-eval verdict exists" pre-ramp gate. check-routing-health owns post-ship live diagnostics. optimize-workload / optimize-agentic-workload own making the workload better — this gate only judges a change, it never optimizes.

Safety Gates

  • Read-only toward production. No route writes, no traffic changes, no capture-setting changes. The only side effects are local artifacts under .understudy/simulate-before-launch/ and the inference calls of the replay itself.
  • Name the spend before running. The gate replays N rollouts × M tasks × 2 arms through a serving path that may bill. State the rollout count and a cost estimate, and get approval when the run is not local-only.
  • No verdict on stale artifacts. The verdict binds to the harness_sha256 / metric_sha256 / splits_sha256 triple from capture-evidence. If any hash no longer matches, refuse and route to re-baseline — a verdict against a moved target is noise wearing a badge.
  • Synthetic examples only in public. Verdicts and replay rows built from real captures stay local; anything committed uses synthetic fixtures.

Read the full file on GitHub · 151 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 151 lines · 0 tokens per session scan A 35716e299806

Subscribe to this mod's changes

simulate-before-launch is a skill published in the GitHub repository understudylabs/understudy-agent-tools (16 stars, last pushed 3d ago), licensed MIT. It adds 78 tokens to every session and 2,083 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens