rollback-guard

rollback-guard is a skill for Claude Code from iampawan/Manifest. It costs 88 tokens per session (2,019 once invoked), scanned A, original, MIT.

A rollout monitor for a feature moving from a canary release to all users. It checks error reports, crash-free rate, and how many users adopt the release against agreed limits.

In plain words
What is it for?
Use it during a rollout to review Sentry errors, crash rates, and release adoption. Run it automatically on a schedule or manually with a contract ID.
Why use it?
It helps catch a release that becomes unhealthy after deployment, before the problem spreads further. It recommends whether to continue, pause, or roll back, while leaving the final action to a person.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the manifest plugin — 28 skills, 15 commands shipped together

Good fit Use it during a rollout to review Sentry errors, crash rates, and release adoption. Run it automatically on a schedule or manually with a contract ID.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/iampawan/manifest/rollback-guard
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.

Any agent
npx skills add iampawan/Manifest --skill rollback-guard
Clone the repo
git clone --depth 1 https://github.com/iampawan/Manifest

Made for: Claude Code.

Or install manifest, the plugin that ships this one along with the rest of its 28 skills, 15 commands.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for rollback-guard

README.md
[![agentmods](https://agentmods.dev/badge/skills/iampawan/manifest/rollback-guard.svg)](https://agentmods.dev/skills/iampawan/manifest/rollback-guard)
Your own site
<a href="https://agentmods.dev/skills/iampawan/manifest/rollback-guard"><img src="https://agentmods.dev/badge/skills/iampawan/manifest/rollback-guard.svg" alt="Measured on agentmods" height="20"></a>
Per session 88 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,019 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00088 $0.02019
Opus 5 $0.00044 $0.01009
Sonnet 5 $0.00018 $0.00404
Haiku 4.5 $0.00009 $0.00202

Measured 8d ago against content hash 19ae3ab45d55, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

rollback-guard 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 8d 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/rollback-guard/SKILL.md · 185 lines

How it starts

The opening of the file, as written. The whole thing — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Rollback guard

verify-deployment decides whether a build is ready to start rolling out. You are the watcher for the period after that — while the feature ramps from canary to 100%, you keep asking "is this still healthy, or is it quietly on fire?" and surface a clear recommendation the moment it stops being healthy.

You recommend; you do not act. Pausing a canary or rolling back is a production change with blast radius — a human approves and performs it (or flips the flag). Auto-executing a rollback is out of scope by design, and matches verify-deployment's "don't auto-rollback" stance.

Inputs

  • Contract ID (+ revision)
  • Optional: explicit time window (default: since canaryStartedAt, or the last 30 min if a tighter sample is requested)

When you should run

While a contract is in its rollout window: canaryStartedAt is set and prodRollout100At is not — or it just reached 100% and you're in the first 24h of full traffic. Outside that window, the slower launch-report cadence (day 1/7/14/28) takes over. The rollback-guard.yml workflow runs you on a tight cadence during that window; /rollback-check <ID> triggers you manually.

Process

1. Read the contract's budgets

From .manifest/contracts/<ID>.r<N>.md, pull the thresholds you'll judge against. These come from existing fields plus an optional rollbackTriggers block:

  • Per-behavior perfBudget (latency / render / cold-start numbers).
  • The crash-free / error-rate budget if declared.
  • Optional rollbackTriggers frontmatter (explicit guardrails):
rollbackTriggers:
  errorRateMultiplier: 3      # error rate > 3× pre-launch baseline → breach
  crashFreeFloor: 99.0        # crash-free users % must stay ≥ this
  newIssueSeverity: blocker   # any new Sentry issue at/above this → breach
  minAdoptionForSignal: 0.05  # ignore noise until ≥5% of users are on the release

If rollbackTriggers is absent, fall back to sensible defaults (error rate > 3× baseline, crash-free < 99%, any new high-volume unhandled-exception issue), and say in the report that defaults were used.

Read the full file on GitHub · 185 lines

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. 8d ago First seen · 185 lines · 88 tokens per session scan A 19ae3ab45d55

Subscribe to this mod's changes

rollback-guard is a skill published in the GitHub repository iampawan/Manifest (5 stars, last pushed 6d ago), licensed MIT. It adds 88 tokens to every session and 2,019 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 2026-08-31.

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

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 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

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

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens