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 agentmods add instructions/screen-arts/mediawork-mcp/agents-mdgit clone --depth 1 https://github.com/screen-arts/mediawork-mcpWrote 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/instructions/screen-arts/mediawork-mcp/agents-md)<a href="https://agentmods.dev/instructions/screen-arts/mediawork-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/screen-arts/mediawork-mcp/agents-md.svg" alt="Measured on agentmods" 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 | $0.02267 | $0.02267 |
| Opus 5 | $0.01133 | $0.01133 |
| Sonnet 5 | $0.00453 | $0.00453 |
| Haiku 4.5 | $0.00227 | $0.00227 |
Grade A, and why
mediawork-mcp AGENTS.md 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 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -X POST http://localhost:3004/mcp \ How it starts
The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
mcp.mediawork.io for Agents
The Mediawork MCP server: read-only Model Context Protocol access to everything the product publishes publicly — the vendor directory, the FAQ, the blog, and subscription plans.
Connector URL: https://mcp.mediawork.io/mcp (Streamable HTTP, no authentication).
Sibling repos:
screen-arts/mediawork-appis the product app that owns the/api/wwwcontract this server wraps, andscreen-arts/mediawork-wwwis the marketing site — the other consumer of that same contract. Neither is a dependency: this repo talks to the app over public HTTP only.
The two rules that define this project
1. Read-only, structurally. This deployment holds one environment variable, APP_API_URL. No
database credentials, no secrets, no write path to anything. Do not add a tool that mutates, and do
not add a credential that would make one possible — the guarantee is the deployment's shape, not a
promise in a code comment.
2. It only ever wraps the app's public API. Every byte served here comes from
https://app.mediawork.io/api/www/*, the same published contract the marketing site consumes. If a
tool needs data that endpoint doesn't expose, add it to the app's /api/www surface first — never
reach around into a database or a private route.
Layout
The Next.js deployable lives in deploy/; server.json sits at the root because it describes the
listing, not the deployable.
server.json the MCP Registry manifest (see below)
deploy/src/lib/fetch.ts APP_API_URL, cache tags, the one fetch wrapper
deploy/src/lib/directory-api.ts directory endpoints + mirrored response types
deploy/src/lib/content-api.ts FAQ, blog and products endpoints + types
deploy/src/lib/links.ts canonical www.mediawork.io URL builders
deploy/src/mcp/shape.ts pure payload -> tool result shapers
deploy/src/mcp/resolve.ts the `kind:key` id scheme for search/fetch
deploy/src/mcp/tools/*.ts tool registration, one file per group
deploy/src/app/mcp/route.ts the handler, mounted at the endpoint path
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.
- 3d ago First seen · 168 lines · 2,267 tokens per session scan A c7d80cf2f377
mediawork-mcp AGENTS.md is an instructions file published in the GitHub repository screen-arts/mediawork-mcp (0 stars, last pushed 7d ago), licensed MIT. It adds 2,267 tokens to every session, about $0.0113 per session on Opus 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 2026-08-31.
Other instructions, from other repositories
gsd-task-manager copilot-instructions.md
Copilot instructions for vscarpenter/gsd-task-manager, covering copilot instructions for gsd task manager, architecture overview, core stack, database schema (v13) and key architectural decisions.
gsd-task-manager CLAUDE.md
Claude Code instructions for vscarpenter/gsd-task-manager, covering claude.md, project overview, design context, core commands and development.
gsd-task-manager AGENTS.md
AGENTS.md instructions for vscarpenter/gsd-task-manager, covering agents.md, cursor cloud specific instructions, security review surfaces and this is not the next.js you know.
lastest CLAUDE.md
Claude Code instructions for las-team/lastest, covering claude.md, critical rules, commands, tesseract. part of the default compose stack; set and ocrserviceurl=http://localhost:8891 in .env.local.
sp500-mcp-server AGENTS.md
AGENTS.md instructions for zhensherlock/sp500-mcp-server, covering agents.md - sp500-mcp-server, repo shape, commands, runtime and test gotchas and mcp tool wiring.
sp500-mcp-server CLAUDE.md
Claude Code instructions for zhensherlock/sp500-mcp-server: Use AGENTS.md as the primary project instruction file.