webmcpify AGENTS.md

An instruction guide for an AI coding agent that helps make a software project usable through WebMCP, a way for web applications to expose actions to AI agents.

In plain words
What is it for?
Making a project agent-ready with WebMCP, reviewing available routes and interactive elements, integrating tools, verifying them, and resuming unfinished work.
Why use it?
It tells the agent which guide to follow and how to track discovery, approval, integration, testing, repair, and reporting without losing progress.

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/tuejon/webmcpify/agents-md
Clone the repo
git clone --depth 1 https://github.com/TueJon/webmcpify

Made for: Codex, OpenCode.

Per session 648 This file is loaded in full into every session.
When invoked 648 The same file — it is already loaded in full.
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 $0.00648 $0.00648
Opus 5 $0.00324 $0.00324
Sonnet 5 $0.00130 $0.00130
Haiku 4.5 $0.00065 $0.00065

Measured 2d ago against content hash 8e7757736fa9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

webmcpify AGENTS.md 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 2d 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.

AGENTS.md · 45 lines

What it actually says

webmcpify — for agents reading this repo

This repository IS an agent skill. If a user asked you to "webmcpify" a project or make it agent-ready with WebMCP, do not explore this repo further — read and follow skills/webmcpify/SKILL.md inside the target project's working directory.

Quick facts:

  • The pipeline: DETECT → INVENTORY (loop) → human manifest approval → INTEGRATE (loop) → VERIFY (loop) → HEAL (loop) → AUDIT + report. Invocation modes (inventory / integrate / verify / status / full) run subsets — see SKILL.md §Invocation modes.
  • All state persists in .webmcpify/manifest.json (Manifest v4) in the target repo — resume from it if it exists. Terminal tool statuses: verified, skipped, rejected.
  • Coverage is explicit: curated produces a reviewed route→tool map; parity produces a per-route interactive-element census. Neither mode may infer coverage from a raw tool count.
  • skills/webmcpify/templates/ contains everything you vendor into target projects (runtime TS + JS, ambient types, Playwright spec) — the skill directory is self-contained; nothing outside it is needed at run time.
  • Hard rules live in SKILL.md "Ground rules" — zero unrelated changes (baseline- aware), read-only tools first, server as the only trust boundary, spec-shaped API usage, commits opt-in.

When editing this repo itself: keep SKILL.md token-efficient (details belong in references/), keep the runtime dependency-free and the TS/JS variants in sync, keep every distribution manifest (.claude-plugin/, .cursor-plugin/, gemini-extension.json, package.json) at the same version, and re-check API claims against https://webmachinelearning.github.io/webmcp/ and https://developer.chrome.com/docs/ai/webmcp — the surface is in flux.

Two discoverability constraints, both load-bearing:

  • Keep the phrase "WebMCP agent skill" in the SKILL.md description, the package.json/skill.json descriptions and the README <h1>. webmcpify is a coined single token that matches neither webmcp nor agent skill in search, and these strings are exactly what marketplaces and search engines display. Dropping the phrase is why one directory invented "Antigravity agent skill" and another filed the skill as an "MCP Server".
  • The SKILL.md frontmatter is plain YAML: no : inside an unquoted scalar (use an em dash or "that"), or every consumer's parser breaks. The description already contains double quotes, so it cannot simply be wrapped in quotes either.
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. 2d ago First seen · 45 lines · 648 tokens per session scan A 8e7757736fa9

Subscribe to this mod's changes

webmcpify AGENTS.md is an instructions file published in the GitHub repository TueJon/webmcpify (6 stars, last pushed 3d ago), licensed MIT. It adds 648 tokens to every session, about $0.0032 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 2026-08-31.