Borrowing it
Nothing to install: this file belongs to quangdang46/openproxy. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/quangdang46/openproxy/main/.agents/skills/openproxy-rtk/SKILL.mdgit clone --depth 1 https://github.com/quangdang46/openproxyWrote 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/quangdang46/openproxy/openproxy-rtk)<a href="https://agentmods.dev/skills/quangdang46/openproxy/openproxy-rtk"><img src="https://agentmods.dev/badge/skills/quangdang46/openproxy/openproxy-rtk/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/quangdang46/openproxy/openproxy-rtk"><img src="https://agentmods.dev/badge/skills/quangdang46/openproxy/openproxy-rtk.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00141 | $0.04721 |
| Opus 5.5 | $0.00056 | $0.01888 |
| Sonnet 5.5 | $0.00028 | $0.00944 |
| Haiku 4.5 | $0.00014 | $0.00472 |
Grade A, and why
openproxy-rtk scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 29d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sS http://127.0.0.1:4623/health | jq '.status' How it starts
The opening of the file, as written. The whole thing — 343 lines — stays where its author put it; the contents beside it link to each section on GitHub.
openproxy-rtk — token saving / request compression
OpenProxy's RTK stack is a set of savers that run on the request body after translation and before dispatch, so the provider receives a smaller payload than the client sent. It is not a prompt-engineering feature and it does not touch responses — it compresses the inbound context, primarily tool-call results.
This skill covers:
- The saver chain and exactly where it sits in the dispatch pipeline
- Turning RTK (and the adjacent Caveman / Ponytail / Headroom savers) on and off
- The per-request
x-9router-token-saveropt-out header - Which payloads RTK actually compresses, and which it deliberately leaves alone
- Reading the real savings numbers from
GET /api/compression/stats - PXPIPE — the token-saver that is not wired up, and what to use instead
Everything here is non-interactive and safe to run unattended.
0 · Preflight
curl -sS http://127.0.0.1:4623/health | jq '.status'
# → "ok"
export OPENPROXY_API_KEY="$(openproxy --robot server init | jq -r '.data.admin_key.key')"
If the server is not running: openproxy server start --detach --no-open.
If there is no admin key, mint one with openproxy key add <name> --auto.
/api/compression/stats and every /api/pxpipe/* route sit behind the admin/management gate — send
Authorization: Bearer $OPENPROXY_API_KEY (or a dashboard session cookie). A request with no
header may succeed on a loopback dev instance with login disabled, but a wrong key is always a
401 invalid_api_key. Always send the bearer.
1 · Where the savers run
The dispatch pipeline (src/server/api/chat.rs) applies, in this order, after the request has been
translated into the provider's wire format:
- RTK — tool-result compression (
compress_messages) - Headroom — optional external compressor at
headroomUrl - Caveman / Ponytail — system-prompt injection
- Claude tool dedupe
- TTS tool strip
- PXPIPE image-context pass (disabled — see §6)
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.
- 29d ago First seen · 343 lines · 141 tokens per session scan A e3f479330d00
openproxy-rtk is a skill published in the GitHub repository quangdang46/openproxy (29 stars, last pushed 5d ago), licensed MIT. It adds 141 tokens to every session and 4,721 once invoked, about $0.0006 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 2026-09-29.
Other skills, from other repositories
implementing-api-schema-validation-security
Implement API schema validation using OpenAPI specifications and JSON Schema to enforce input/output contracts and prevent injection, data exposure, and mass assignment attacks.
keirouter-chat
Chat / code generation via KeiRouter using OpenAI /v1/chat/completions or Anthropic /v1/messages format with streaming + auto-fallback combos. Use when the user wants to ask an LLM, generate code, summarize text, or run prompts through KeiRouter.
keirouter-embeddings
Generate vector embeddings via KeiRouter /v1/embeddings using OpenAI / Mistral / Cohere / Voyage / NVIDIA / AWS Bedrock and other embedding models for RAG, semantic search, similarity. Use when the user wants embeddings, vectors, RAG, semantic search, or to embed text.
10router
Entry point for 10Router — local/remote AI gateway with OpenAI-compatible REST for chat, image, TTS, embeddings, web search, web fetch. Use when the user mentions 10Router, TENROUTERURL, or wants AI without writing provider boilerplate. This skill covers setup + indexes capability skills; fetch the relevant capability…
post-processing
Extract, analyze, and summarize simulation output data — pull spatial fields at specific timesteps, compute time-series trends and detect steady state, extract line profiles through the domain, generate statistical summaries and distributions, calculate derived quantities (gradients, fluxes, volume fractions…
linear-solvers
Select and configure linear solvers for Ax=b systems arising in numerical simulations — choose between direct (LU, Cholesky) and iterative (CG, GMRES, BiCGSTAB, MINRES) methods, analyze sparsity patterns and matrix conditioning, recommend preconditioners (AMG, ILU, IC), apply row/column scaling, and diagnose…