A guide for finding web cache poisoning and cache deception. Web caches store responses for reuse; poisoning puts attacker-controlled content into that stored response, while deception makes a private response look cacheable.

Use it to analyze CDN or application caches, test reflected headers and parameters, inspect cache keys, and check path or delimiter tricks.

It helps reveal when headers, parameters, or paths affect a response without affecting the cache key, allowing one user's content to reach others.

hunt-cache — Skill for Claude Code by Encod3d-Sec/TORCH

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 Encod3d-Sec/TORCH --skill hunt-cache

Clone the repo

git clone --depth 1 https://github.com/Encod3d-Sec/TORCH

Made for: Claude Code.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/encod3d-sec/torch/hunt-cache/github.svg)](https://agentmods.dev/skills/encod3d-sec/torch/hunt-cache)

Your own site

<a href="https://agentmods.dev/skills/encod3d-sec/torch/hunt-cache"><img src="https://agentmods.dev/badge/skills/encod3d-sec/torch/hunt-cache/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/encod3d-sec/torch/hunt-cache"><img src="https://agentmods.dev/badge/skills/encod3d-sec/torch/hunt-cache.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 39 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,521 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

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.

Model	Per session	Once invoked
Fable 5.1	$0.00039	$0.01521
Opus 5.5	$0.00016	$0.00608
Sonnet 5.5	$0.00008	$0.00304
Haiku 4.5	$0.00004	$0.00152

Measured 1mo ago against content hash 141b2322c199, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

hunt-cache scanned grade A with 1 finding 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 1mo 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.

Makes network callslowCapability

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

**Drive load-bearing requests through Burp Repeater** for operator visibility; use Param Miner to enumerate unkeyed headers/params. curl is fine for the quick keyed-vs-unkeyed loop.

skills/hunt/hunt-cache/SKILL.md · 75 lines

How it starts

The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Hunt: Web Cache Attacks

Assumes hunt-core for the scope gate, two-account rule, confirmation gate, enumeration limits, stop conditions, wiki protocol, FIND output, and Deadends. Do not re-derive any of that here.

Wiki

qmd_query "web cache poisoning deception unkeyed input cache-key path confusion" via wiki-search MCP

Hub: [[web-moc]] (live web index). Primary page: [[web-cache-poisoning]]. Payload arsenal: wiki/payloads/web-cache.md. Anchors: [[web-cache-deception]], [[web-cache-attacks]]. Related: [[http-host-header-attacks]], [[http-request-smuggling]].

Attack surface

Needs a cache in front (CDN / Varnish / Cloudflare / Akamai / Fastly, or an app-layer cache). Signals: Age, X-Cache: hit/miss, Cache-Control, CF-Cache-Status headers; static-ish responses; responses that reflect a header or param.

Rank before testing:

Unkeyed headers that reflect - X-Forwarded-Host, X-Forwarded-Scheme, X-Host, X-Forwarded-For, plus custom headers a page reflects into links/scripts. Highest hit-rate poisoning vector; discover unkeyed inputs with Param Miner.
Path / parameter cloaking - static-looking suffixes and delimiter tricks (/account/profile.css, /account/profile/nonexistent.js, path-parameter ;, encoded %2f, fat GET) that desync what the cache keys on from what the origin serves. Primary deception vector.
CDN edges and normalization gaps - cache-key normalization (case, trailing slash, duplicate params) differing from origin routing; multi-CDN or origin-vs-edge disagreement.

Methodology

Drive load-bearing requests through Burp Repeater for operator visibility; use Param Miner to enumerate unkeyed headers/params. curl is fine for the quick keyed-vs-unkeyed loop.

Identify the cache + cache key. Compare X-Cache/Age across requests; determine what is keyed (usually method + host + path + some query) vs unkeyed (most headers, some params). Always attach a unique cache-buster while probing so you never touch a shared key.
Cache poisoning (unkeyed input -> harmful response, then cached for others).
- Find an unkeyed input that affects the response (reflected header/param): X-Forwarded-Host, X-Forwarded-Scheme, X-Host, X-Forwarded-For, custom headers (Param Miner to discover).
- Make it produce harm (XSS / redirect / resource swap), then confirm the cached poisoned response is served to a fresh request (cache-buster off) - and cross-session (see confirmation gate).
- Fat GET, parameter cloaking, and cache-key normalization gaps as variants.
Cache deception (trick the cache into storing a victim's private page).
- Request a private page with an appended static-looking suffix/path: /account/profile.css, /account/profile/nonexistent.js, path-parameter ;, encoded %2f.
- If the origin returns the private content but the cache stores it as static -> retrieve another user's data unauthenticated.
Confirm impact crosses a trust boundary - served to other users / discloses private data, not just your own session.
Distill when confirmed (reusable unkeyed-header or deception-path trick, GENERIC, no client host): python3 scripts/wiki-stage.py --kind technique --slug <slug> --target-page techniques/web/web-cache-poisoning.md (deception-path findings: --target-page techniques/web/web-cache-deception.md).

Read the full file on GitHub · 75 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.

1mo ago First seen · 75 lines · 39 tokens per session scan A 141b2322c199

Subscribe to this mod's changes

hunt-cache is a skill published in the GitHub repository Encod3d-Sec/TORCH (336 stars, last pushed 1mo ago), licensed MIT. It adds 39 tokens to every session and 1,521 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Backend and APIs Security

mem0-oss-to-platform

Plan, then execute, a migration of a project from the mem0 open-source / self-hosted SDK (local Memory class) to the mem0 Platform / hosted SDK (MemoryClient). TRIGGER when: the developer wants to move, switch, or migrate mem0 off OSS/self-hosted to the hosted API, e.g. "migrate my mem0 setup to the platform", "switch…

mem0ai/mem0 · 217 tokens

agui-dotnet-protobuf

Use the protobuf wire transport (instead of the default Server-Sent Events) for an AG-UI connection with the AG-UI .NET SDK — a compact binary event stream negotiated via the Accept header. USE FOR: making an AGUIChatClient prefer protobuf by wiring an AGUIEventStreamHandler with ProtobufEventStreamFormatter (then…

ag-ui-protocol/ag-ui · 162 tokens

azure-mgmt-botservice-dotnet

Azure Resource Manager SDK for Bot Service in .NET. Management plane operations for creating and managing Azure Bot resources, channels (Teams, DirectLine, Slack), and connection settings. Triggers: "Bot Service", "BotResource", "Azure Bot", "DirectLine channel", "Teams channel", "bot management .NET", "create bot".

microsoft/skills · 78 tokens

fastapi-router-py

Create FastAPI routers with CRUD operations, authentication dependencies, and proper response models. Use when building REST API endpoints, creating new routes, implementing CRUD operations, or adding authenticated endpoints in FastAPI applications.

microsoft/skills · 46 tokens

dev-runtime

How the Node side of a dev stack runs without haven, and the backend host: pnpm dev (UI built once, backend reloading), pnpm dev:hmr (adds the Vite ui lane), tools/dev-runtime (api + worker in one Node process), backend reload on a file change, the debounce (quiet window + max wait), ports derived from PORT. Use when…

langwatch/langwatch · 172 tokens

migrate-segw-to-rap

Reverse-engineer a SEGW-built OData V2 service (MPC/DPC/MPCEXT/DPCEXT) into a modern RAP V4 service — tables, CDS views (interface + projection), behavior definitions, draft entities, service definition + binding. Use when asked to "migrate this SEGW service to RAP", "convert OData V2 to V4 RAP", "modernize this…

arc-mcp/arc-1 · 106 tokens

hunt-cache