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.
npx skills add forefy/.context --skill http-request-smugglinggit clone --depth 1 https://github.com/forefy/.contextWrote 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.
[](https://agentmods.dev/skills/forefy/.context/http-request-smuggling)<a href="https://agentmods.dev/skills/forefy/.context/http-request-smuggling"><img src="https://agentmods.dev/badge/skills/forefy/.context/http-request-smuggling/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.
<a href="https://agentmods.dev/skills/forefy/.context/http-request-smuggling"><img src="https://agentmods.dev/badge/skills/forefy/.context/http-request-smuggling.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00046 | $0.02470 |
| Opus 5.5 | $0.00018 | $0.00988 |
| Sonnet 5.5 | $0.00009 | $0.00494 |
| Haiku 4.5 | $0.00005 | $0.00247 |
Grade A, and why
http-request-smuggling 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.
Everything below sends bytes on a raw socket - you cannot craft these requests with `curl`, because clients and libraries normalize the conflicting headers and fix the framing. Send the exact bytes, including CRLF line e How it starts
The opening of the file, as written. The whole thing — 204 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Contents
- Scope & authorization
- Why raw sockets
- Methodology (timing probes then differential confirmation)
- Payloads (CL.TE and TE.CL, exact bytes)
- Runnable snippet
- Output
Scope & authorization
Request smuggling is an ACTIVE, potentially disruptive check. A successful desync poisons the shared front-end-to-back-end connection: the smuggled prefix can attach to the next request on that socket, which in a real deployment may belong to another user. Run this only against hosts you own or are contractually engaged to test, and prefer a low-traffic window. The timing probes (Steps 2-3) are the safest first pass because they never leave a dangling prefix; the differential confirmations (Steps 4-5) do smuggle a real prefix and should follow only once timing suggests a desync.
The vulnerability exists when a request traverses two servers (CDN/load-balancer/reverse-proxy front-end, then an origin back-end) that disagree on where one request ends and the next begins. That disagreement comes from sending both a Content-Length (CL) and a Transfer-Encoding: chunked (TE) header: if the front-end honors one and the back-end the other, the leftover bytes are "smuggled." A single server with no proxy in front cannot desync, so a negative result on a direct-to-origin host is expected, not a clean bill.
Everything below sends bytes on a raw socket - you cannot craft these requests with curl, because clients and libraries normalize the conflicting headers and fix the framing. Send the exact bytes, including CRLF line endings.
Why raw sockets
- The two framing headers (
Content-LengthandTransfer-Encoding) must both be present and must reach the server unmodified. HTTP libraries strip or reconcile them. - Line endings must be real CRLF (
\r\n). Author the payload with plain\nfor readability, then convert every\nto\r\nbefore sending - matching the source templates exactly. - Detection is per-connection and stateful: you open a fresh TLS socket, send one crafted request, read the response, and compare across two identical sends. The desync shows up on the second request.
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 · 204 lines · 46 tokens per session scan A 51333e12c081
http-request-smuggling is a skill published in the GitHub repository forefy/.context (153 stars, last pushed yesterday), licensed MIT. It adds 46 tokens to every session and 2,470 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.
Other skills, from other repositories
fizz
Generate Echidna/Medusa-compatible Solidity fuzz suites from Foundry or Hardhat projects. Trigger on "fizz", "generate fuzz suite", "build fuzz harness", "stateful fuzzing", "fuzzing harness", "property testing", and "invariant suite".
asc-shots-pipeline
Orchestrate iOS screenshot automation with xcodebuild/simctl for build-run, AXe for UI actions, JSON settings and plan files, Koubou-based framing (asc screenshots frame), and screenshot upload (asc screenshots upload). Use when users ask for automated screenshot capture, AXe-driven simulator flows, frame composition…
terraform-skill
Terraform infrastructure as code best practices.
endpoint-validator
Deterministic API endpoint validation with pass/fail reporting.
qa-manual-istqb
Create QA artifacts from requirements: test plans, test conditions/cases, bug reports, regression suites, traceability, and exploratory charters. Use for test planning, test design, defects, coverage, or QA deliverables. Applies ISTQB risk-based techniques and loads templates only when needed. Keywords: test plan…
webapp-selenium-testing
Author and maintain versioned Selenium WebDriver tests with Java and JUnit 5. Use for creating, debugging, or running Selenium specs, implementing Page Objects, handling explicit waits, capturing screenshots, or setting up Maven test projects. Supports Chrome, Firefox, and Edge. Keywords: Selenium WebDriver, Java…