implement-plugin

A workflow for creating Starlark plugins, which are small programs written in the Starlark language, for a proxy's plugin engine. It guides protocol, event, and execution-phase choices before generating and testing a plugin.

In plain words
What is it for?
Scaffolding plugins for HTTP, WebSocket, gRPC, SSE, TLS, SOCKS5, and other supported protocols; selecting events and phases; then running sample tests.
Why use it?
It turns the plugin's technical setup into a repeatable process and helps avoid choosing an invalid hook or configuration.

Skill for Claude CodeCodex

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/usk6666/yorishiro-proxy/implement-plugin
Any agent
npx skills add usk6666/yorishiro-proxy --skill implement-plugin
Clone the repo
git clone --depth 1 https://github.com/usk6666/yorishiro-proxy

Made for: Claude Code, Codex.

Per session 22 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,698 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 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 $0.00022 $0.02698
Opus 5 $0.00011 $0.01349
Sonnet 5 $0.00004 $0.00540
Haiku 4.5 $0.00002 $0.00270

Measured 3d ago against content hash 6fa28e576162, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

implement-plugin 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 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.

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.

.claude/skills/implement-plugin/SKILL.md · 151 lines

How it starts

The opening of the file, as written. The whole thing — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/implement-plugin

A workflow skill for creating Starlark plugins on the RFC-001 §9.3 plugin engine (internal/pluginv2/). Interactively select protocol/event/phase to generate a working scaffold.

Arguments

  • /implement-plugin — Create a plugin in interactive mode
  • /implement-plugin <description> — Auto-generate a plugin from a description (e.g., /implement-plugin Add X-Request-ID header to HTTP requests)

Steps

Phase 1: Requirements Clarification

If no description argument is given, interactively confirm:

  1. Purpose: What should the plugin do?
  2. Protocol: Target plugin protocol — one of http, ws, grpc, grpc-web, sse, raw, tls, connection, socks5 (the plugin protocol vocabulary; see internal/pluginv2/surface.go).
  3. Event: Hook event under that protocol (e.g. on_request, on_message, on_chunk).
  4. Phase: pre_pipeline (default — before Intercept/Transform) or post_pipeline (after; right place for signing/last-mile mutation). Lifecycle / observation events (on_close, on_handshake, on_disconnect, socks5.on_connect, etc.) take no phase= argument — passing one is a load-time error.
  5. Action: CONTINUE (mutate in place) / DROP (transaction-start only) / RESPOND (transaction-start request, plus http.on_response for replacement).

If a description argument is given, auto-determine from its content.

Phase 2: Context Collection

Read the following to understand the plugin API:

  1. docs/rfc/plugin-migration.md — Direct migration table from legacy hook names to (protocol, event, phase) and the rationale.
  2. internal/pluginv2/surface.go — The 17-entry hook surface table. Authoritative for which (protocol, event) pairs exist and which actions each accepts.
  3. internal/pluginv2/convert.go — The Starlark msg dict shape per protocol (snake_case field names, *HeadersValue ordered list, msg["raw"] first-class byte injection).
  4. internal/pluginv2/lifecycle.go — Dict shapes for lifecycle events (connection.on_connect, socks5.on_connect, tls.on_handshake, ws.on_close).
  5. internal/pluginv2/respond_action.go — Typed action.RESPOND(...) / action.RESPOND_GRPC(...) builtin signatures.
  6. examples/plugins/ — Existing sample plugins for each major protocol.

Read the full file on GitHub · 151 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. 3d ago First seen · 151 lines · 22 tokens per session scan A 6fa28e576162

Subscribe to this mod's changes

implement-plugin is a skill published in the GitHub repository usk6666/yorishiro-proxy (16 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 22 tokens to every session and 2,698 once invoked, about $0.0001 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 2026-08-30.

Related

Other skills, from other repositories

screenshots

Advanced UI analysis and interaction through pixel-level reasoning to solve complex interaction problems, locate hard-to-find elements, verify visual states, and debug layout/rendering issues. Use as needed after UI Mapper inspection when advanced reasoning or non-standard views are required.

xAmirHamza77/ReverseOps-Skill · 52 tokens

ReverseOps-router

Routes reverse engineering, exploitation, penetration testing, malware, mobile, firmware, browser automation, documentation, and security tasks to the appropriate specialist skill. Use when a task spans modules or the correct ReverseOps entrypoint is unclear.

xAmirHamza77/ReverseOps-Skill · 49 tokens

apk-reverse

Use when performing Android APK reverse engineering in a CLI environment. Applies to APK unpacking, Java decompilation, smali modification, repackaging, and Frida dynamic hooking, with on-demand switching to so/native analysis. Prefer locally installed jadx, apktool, frida, adb, ida-reverse, radare2.

xAmirHamza77/ReverseOps-Skill · 71 tokens

binary-diff

Cross-version symbol migration and binary diffing. Use this when you have symbols/reverse engineering results from an old version and need to quickly migrate them to a new version. Applicable scenarios: kernel missing PDB with derivation from old-version symbols, batch-migrating function names after a program update…

xAmirHamza77/ReverseOps-Skill · 129 tokens

dotnet-reverse

.NET / C# binary reverse engineering. Use when the target is a .NET assembly (CLR in the PE header, managed .exe/.dll programs), C# build artifacts (including NativeAOT), red team Sharp tools (Rubeus / SharpHound, etc.), .NET obfuscated programs (ConfuserEx / SmartAssembly / Babel / Eazfuscator), or .NET loaders /…

xAmirHamza77/ReverseOps-Skill · 132 tokens

edr-bypass-re

Reverse engineering defender implementations → Red Team targeted bypass. Reverse engineer EDR / Defender / AV hook tables, ETW providers, and AMSI implementations first, then write targeted unhooking / indirect syscalls / ETW patches / call stack spoofing. Aligned with MITRE ATT&CK T1562 Impair Defenses. Trigger…

xAmirHamza77/ReverseOps-Skill · 175 tokens