cpv-cache-validation-skill

cpv-cache-validation-skill is a skill for Claude Code, Codex from Emasoft/claude-plugins-validation. It costs 65 tokens per session (1,413 once invoked), scanned B, original, MIT.

A checker for Claude Code plugins and projects that looks for seven documented prompt-cache invalidation patterns. Prompt caching reuses unchanged instructions to reduce repeated processing, so changing any part of the cached prefix can cause a cache miss.

In plain words
What is it for?
Use it when auditing a plugin or project for cache regressions, or when fixing CA-01 through CA-07 findings.
Why use it?
It helps find configuration and file patterns that unexpectedly break caching or create many separate cache versions. The findings are warnings because they affect cost and delay, not whether the plugin works.

Skill for Claude CodeCodex

Installs and runs on its own, but its text points at files inside its plugin — anything it tells you to read at a ${CLAUDE_PLUGIN_ROOT} path is only there once the plugin is installed. Installing the plugin gets both.

Part of the claude-plugins-validation plugin — 31 skills, 14 commands, 14 agents shipped together

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/emasoft/claude-plugins-validation/cpv-cache-validation-skill
Any agent
npx skills add Emasoft/claude-plugins-validation --skill cpv-cache-validation-skill
Clone the repo
git clone --depth 1 https://github.com/Emasoft/claude-plugins-validation

Made for: Claude Code, Codex.

Or install claude-plugins-validation, the plugin that ships this one along with the rest of its 31 skills, 14 commands, 14 agents.

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 cpv-cache-validation-skill

README.md
[![agentmods](https://agentmods.dev/badge/skills/emasoft/claude-plugins-validation/cpv-cache-validation-skill.svg)](https://agentmods.dev/skills/emasoft/claude-plugins-validation/cpv-cache-validation-skill)
Your own site
<a href="https://agentmods.dev/skills/emasoft/claude-plugins-validation/cpv-cache-validation-skill"><img src="https://agentmods.dev/badge/skills/emasoft/claude-plugins-validation/cpv-cache-validation-skill.svg" alt="Measured on agentmods" height="20"></a>
Per session 65 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,413 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 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 $0.00065 $0.01413
Opus 5 $0.00032 $0.00707
Sonnet 5 $0.00013 $0.00283
Haiku 4.5 $0.00006 $0.00141

Measured 3d ago against content hash 87192047b69f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

cpv-cache-validation-skill scanned grade B with 2 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 3d 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.

Downloads and executes remote codemediumSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

runs. Install via `curl -LsSf https://astral.sh/uv/install.sh | sh`.

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

runs. Install via `curl -LsSf https://astral.sh/uv/install.sh | sh`.
skills/cpv-cache-validation-skill/SKILL.md · 100 lines

How it starts

The opening of the file, as written. The whole thing — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cache-Audit Skill (loaded by cpv-cache-optimizer-agent)

Overview

Anthropic's prompt cache caches the rendered system-prompt prefix — CLAUDE.md, cached agent/skill bodies, and settings-derived blocks. The cache is invalidated whenever ANY byte in that prefix changes. A 200K-token prefix cache MISS costs ~10x normal token rate vs a HIT. CPV's cache-audit rule pack catches the seven documented patterns that silently break caching or fork the cached prompt into many distinct keys.

Severity: every CA-01..CA-07 finding is a WARNING (non-blocking — a cache miss costs tokens/latency but never makes a plugin invalid). validate_plugin CALLS this scanner as a SEPARATE step (its own report + a one-line pointer in the main report); this skill + cpv-cache-optimize act on the findings. CA-04 covers a model: frontmatter on ANY component (agents, commands, skills); model: inherit is exempt.

Per-rule severity, catch description, and fix-recipe pointer live in references/ca-rules.md. Full TOC of every rule is embedded under "Resources" at the bottom of this file.

Prerequisites

  • A Claude Code plugin directory OR a project root that uses Claude Code.
  • uv available on PATH so the validator can run via the launcher (uv run --with pyyaml python "${CLAUDE_PLUGIN_ROOT}/scripts/remote_validation.py" cache <path>).

Scanner contract

Same I/O contract as validate_security.py: path-only stdout (auto-save report, print compact summary), aggregated (level, rule_id) grouping, self-scan filter chain (skips catalog / test / dev-scratch files).

Instructions

  1. Read references/launcher-invocation.md for the canonical bash one-liner (alias cache).

    The one-liner · Why the launcher is mandatory · Direct invocation (development only)

  2. Resolve MAIN_ROOT via git worktree list | head -n1.
  3. Run the launcher with --report "$MAIN_ROOT/reports/validate_cache/<TS>-<slug>.md".
  4. Read the compact summary from stdout, full details from the report file.
  5. (Fix workflow only — cpv-cache-optimizer-agent.) Group findings by CA-NN, apply cache-fixes.md#ca-nn from skills/cpv-fix-validation/references/, re-run, iterate until ZERO CA findings remain. Every CA finding is a WARNING, so the verdict is VALID from the start — terminate on an empty findings set, not on VALID.

Read the full file on GitHub · 100 lines

Files

What ships with it

2 files 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. 3d ago First seen · 100 lines · 65 tokens per session scan B 87192047b69f

Subscribe to this mod's changes

cpv-cache-validation-skill is a skill published in the GitHub repository Emasoft/claude-plugins-validation (4 stars, last pushed yesterday), licensed MIT. It adds 65 tokens to every session and 1,413 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 2 findings (downloads and executes remote code, makes network calls). 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

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

auto-perf-optimize

Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.

microsoft/vscode · 62 tokens

chat-perf

Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.

microsoft/vscode · 51 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 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