actionproxy AGENTS.md

Súbor pravidiel pre prácu s verejným repozitárom ActionProxy, ktorý riadi a zaznamenáva volania nástrojov AI agentov.

In plain words
What is it for?
Je určený na orientáciu pri úpravách, prijímaní zmien z iného repozitára a dodržiavaní architektúry, bezpečnostných pravidiel a pravidiel API.
Why use it?
Pomáha agentovi pochopiť hranice projektu, správny postup pred úpravami a dokumenty, ktoré si má prečítať.

Instructions file for CodexOpenCode

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 instructions/jandybottleshaped403/actionproxy/agents-md
Clone the repo
git clone --depth 1 https://github.com/Jandybottleshaped403/actionproxy

Made for: Codex, OpenCode.

Per session 1,898 This file is loaded in full into every session.
When invoked 1,898 The same file — it is already loaded in full.
Security scan C 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 $0.01898 $0.01898
Opus 5 $0.00949 $0.00949
Sonnet 5 $0.00380 $0.00380
Haiku 4.5 $0.00190 $0.00190

Measured yesterday against content hash 1ae5c160fc4f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

actionproxy AGENTS.md scanned grade C 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 yesterday.

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.

Hidden instructionshighPrompt injection

Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.

<!-- actionproxy-public-agent-instructions:v1 -->

Makes network callslowCapability

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

`examples/local-curl-demo/` or state why that validation was not run.
AGENTS.md · 200 lines

How it starts

The opening of the file, as written. The whole thing — 200 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ActionProxy Community Repository Instructions

These instructions apply to the whole public repository. Follow explicit user instructions first, then the rules in this file.

Product boundary

ActionProxy Community is an execution-governance gateway for AI agent tool calls. It receives a proposed call, evaluates deterministic policy, allows, denies, or pauses it for approval, executes only with valid authority, and records lifecycle evidence.

Keep the product narrow. It is not an agent runtime, chatbot, browser, generic workflow builder, connector marketplace, hosted control plane, or production SaaS connector bundle.

Read before editing

Start with README.md and CONTRIBUTING.md. Read the documents relevant to the change:

Adopting from another repository

When the user wants to add ActionProxy to an existing application, read the third-party adoption guide before changing either repository. Do not treat adoption as an ActionProxy contribution.

Choose exactly one execution boundary: the MCP wrapper for an existing MCP server, runExternalAction for a JavaScript or TypeScript runner, or the documented HTTP grant lifecycle for another runtime. Begin with a simulated effect and prove zero dispatch before approval, exactly one after approval, zero after denial, and a valid audit chain. ActionProxy v0.1 packages are not published to npm; use the guide's explicit local-tarball workflow rather than inventing a registry dependency or importing across repository boundaries.

Repository map

  • apps/server/src/app.ts composes the Community Fastify application.
  • apps/server/src/services/action-gate.ts owns the governed lifecycle.
  • apps/server/src/policy/ contains pure policy loading and evaluation.
  • apps/server/src/storage/ contains storage interfaces and adapters.
  • apps/server/src/routes/ validates transport input and maps service errors.
  • apps/server/src/tools/ contains deterministic local mock tools.
  • apps/web/src/ contains the local operator console.
  • packages/sdk-js/ contains the JavaScript client and runner helpers.
  • packages/mcp-wrapper/ contains the stdio MCP proxy.
  • examples/google-workspace-mcp-demo/ is the opt-in real-provider reference; its third-party MCP process, not ActionProxy, owns Google OAuth.
  • examples/ contains local demonstrations; docs/ is the contract source of truth.

Read the full file on GitHub · 200 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. yesterday First seen · 200 lines · 1,898 tokens per session scan C 1ae5c160fc4f

Subscribe to this mod's changes

actionproxy AGENTS.md is an instructions file published in the GitHub repository Jandybottleshaped403/actionproxy (0 stars, last pushed yesterday), licensed Apache-2.0. It adds 1,898 tokens to every session, about $0.0095 per session on Opus 5. A static security scan graded it C with 2 findings (hidden instructions, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.