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/buildinternet/uploads/agents-mdgit clone --depth 1 https://github.com/buildinternet/uploadsWhat 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 | $0.04167 | $0.04167 |
| Opus 5 | $0.02083 | $0.02083 |
| Sonnet 5 | $0.00833 | $0.00833 |
| Haiku 4.5 | $0.00417 | $0.00417 |
Grade A, and why
uploads 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 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.
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.
How it starts
The opening of the file, as written. The whole thing — 294 lines — stays where its author put it; the contents beside it link to each section on GitHub.
uploads
File-hosting backend for uploads.sh. Provider-agnostic storage via files-sdk, deployed to Cloudflare Workers with Wrangler. The service is public: anyone can sign up, create a workspace, and use the CLI, and paid plans are live.
Layout
The path-by-path inventory lives in the README's
What's in this repo table — one copy, so it
can't drift. packages/uploads also ships uploads mcp, a stdio MCP server
mirroring the CLI commands.
Three agent skills are checked in at the repo root so they're installable via
the npx skills add convention (and by uploads install):
skills/github-screenshots is the thin workflow skill (when a screenshot or
recording should go into a PR/issue or be shared as a link — the in-repo
successor to the external github-screenshots skill's bundled R2 scripts),
skills/annotate-screenshots covers callouts and redaction
(uploads annotate / screenshot --annotate), and skills/uploads-cli is
the full CLI reference the others defer to. Keep all three in sync when the
CLI's commands or flags change.
Screenshots: stage as you go. If your change is visually observable (web
UI, email templates, rendered output), capture and stage screenshots at each
milestone while you work — don't wait for a PR to exist. A bare uploads put
already stages by default on a non-default git branch (issue #403), and a
bare uploads screenshot (no --pr/--issue/--branch) does the same
(issue #469) — carrying its derived metadata (path/url/env/viewport,
plus --state) through to the PR once it opens; reach for attach --branch
when you want its extras (multiple files in one call, or triggering
promotion/comment sync as a side effect):
uploads put ./after.png --meta path=/settings --state after # before|after|empty|error|loading
uploads screenshot http://localhost:4321/settings --out after.png --state after
uploads attach ./after.png --branch --state after # explicit form, either way
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.
- yesterday First seen · 294 lines · 4,167 tokens per session scan A c958f4554b72
uploads AGENTS.md is an instructions file published in the GitHub repository buildinternet/uploads (19 stars, last pushed yesterday), licensed Apache-2.0. It adds 4,167 tokens to every session, about $0.0208 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.
Other instructions, from other repositories
agents AGENTS.md
Instructions for cloudflare/agents, covering agents.md, project overview, repository structure, nested agents.md files and setup.
mcpu AGENTS.md
Instructions for acoyfellow/mcpu, a project described as: MCP control surface for a Cloudflare Artifacts repo. Bootstrap from GitHub, then edit, commit, and deploy on Cloudflare.
hatch3r CLAUDE.md
Instructions for hatch3r/hatch3r, covering hatch3r — development instructions, architecture, development commands, two-axis pillar framework (2.0.0) and orchestrator self-discipline (bypass protection).
aeon CLAUDE.md
Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
ken CLAUDE.md
Instructions for townsendmerino/ken, covering claude.md, what this is, repository ownership (read this first), commands and embedding parity & golden fixtures (now in aikit).
gangsta GEMINI.md
Instructions for kucherenko/gangsta, a project described as: AI agentic skills framework for spec-driven development, built on the organizational model of mafia.