camelAI AGENTS.md

camelAI AGENTS.md is an instructions file for Codex, OpenCode from qaml-ai/camelAI. It costs 12,380 tokens per session, scanned A, original, MIT.

A project guide for camelAI, a coding-assistant platform built on Cloudflare. It explains how user workspaces, chat sessions, stored files, builds, and published applications fit together.

In plain words
What is it for?
Use it to understand the repository layout and request routing, and to work on persistent coding sessions, project files, sandbox builds, deployments, uploads, and model access.
Why use it?
It helps coding agents navigate a distributed codebase whose parts run in different Cloudflare services. This makes it easier to find the right code when changing workspace storage, agent sessions, builds, or deployment.

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/qaml-ai/camelai/agents-md
Clone the repo
git clone --depth 1 https://github.com/qaml-ai/camelAI

Made for: Codex, OpenCode.

Wrote 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.

agentmods badge for camelAI AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/qaml-ai/camelai/agents-md.svg)](https://agentmods.dev/instructions/qaml-ai/camelai/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/qaml-ai/camelai/agents-md"><img src="https://agentmods.dev/badge/instructions/qaml-ai/camelai/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 12,380 This file is loaded in full into every session.
When invoked 12,380 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.12380 $0.12380
Opus 5 $0.06190 $0.06190
Sonnet 5 $0.02476 $0.02476
Haiku 4.5 $0.01238 $0.01238

Measured yesterday against content hash efa4dd343b7d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

camelAI 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 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.

Makes network callslowCapability

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

curl "https://api.cloudflare.com/client/v4/accounts/$CF_ACCOUNT_ID/analytics_engine/sql" \
AGENTS.md · 397 lines

How it starts

The opening of the file, as written. The whole thing — 397 lines — stays where its author put it; the contents beside it link to each section on GitHub.

camelAI Agent Guide

Keep this file concise and durable. Add details here only when they help future agents navigate the repo or avoid common mistakes. Feature-specific behavior should usually live near the code or in tests.

What This Is

camelAI is an AI coding assistant platform on Cloudflare Workers + Durable Objects with Cloudflare sandbox containers for builds/analysis. Users chat with persistent coding workspaces; each thread runs camelAI's pi-based coding harness in ChatThreadDO, and generated apps publish to *.camelai.app / environment-specific app hosts.

High-Level Architecture

React Router SSR + browser WS
        |
        v
Cloudflare main Worker + Durable Objects
        |
        v
Project files in WorkspaceFilesystemDO + R2 (do-r2 backend)
Builds/deploys + analysis in Cloudflare sandbox containers

Dispatcher Worker routes published user apps.
R2 stores uploads/assets/previews.
Cloudflare AI Gateway and BYOK credentials back model access.

Agent turns run in ChatThreadDO (Pi coding agent). Project source files live in WorkspaceFilesystemDO + R2 (every project is backend: "do-r2"); builds, deploys, and analysis run in Cloudflare sandbox containers (ProjectBuildSandbox, AnalysisSandbox, DbQuerySandbox). The legacy Azure project-runtime-service VM and its PROJECT_RUNTIME_HOST bridge are gone; the only remaining VM is the static-IP database egress relay (infra/db-egress-relay/, see docs/db-egress-relay.md). SQL queries/exports run in DbQuerySandbox (the DATA_PROXY binding is served worker-side). There is no in-repo Go sandbox-host or data-proxy tree.

Repository Map

  • src/ - React Router 7 app, routes, loaders/actions, UI components, shared server/client libraries.
  • src/routes.ts - Imperative React Router route config. Add page/API routes here.
  • src/routes/api/ - React Router API routes for most user-facing REST (billing checkout, workspaces, chat groups, etc.).
  • src/components/ui/ - shadcn/ui components.
  • workers/main/ - Main Cloudflare Worker, Durable Objects, HTTP/SSE transports (plus the log-tail WebSocket /ws/logs), MCP, admin APIs, proxies, container image Dockerfiles.
  • workers/main/src/identity/ - UserDO / OrgDO and related identity helpers (auth.ts is a compatibility barrel).
  • workers/main/src/routes/ - Worker-native HTTP (SSE streams, the log-tail WebSocket /ws/logs, Stripe webhook, data-proxy, MCP, most /api/admin/* on Hono). Prefer documenting new paths here vs src/routes/api/ — see API routing below.
  • workers/dispatcher/ - Workers for Platforms dispatcher for deployed user apps.
  • workers/app-usage-guard/ - Account-wide Durable Object SQLite usage monitor and reversible app quarantine Worker; see docs/deployed-app-usage-guard-design.md.
  • workers/bedrock-provider/ - AI Gateway custom provider translating Anthropic-style requests to Bedrock.
  • workers/user-logs-tail/ - Tail worker for deployed app logs.
  • workers/e2e-reports/ - Public viewer at e2e-reports.camelai.dev serving Playwright E2E reports from R2 (uploaded by the E2E workflow); deploy with bun run deploy:e2e-reports.
  • workers/eval-reports/ - Read-only results store + viewer for agent evals at evals.camelai.dev (evals run locally; EVAL_REPORT=1 publishes them); deploy with bun run deploy:eval-reports.
  • The Go data-proxy (external qaml-ai/project-runtime-service cmd/data-proxy) is retired: SQL queries and warehouse exports now run in the DbQuerySandbox Cloudflare container (workers/main/src/db-query-service.ts + data-proxy.ts compat surface), and the SANDBOX_HOST VPC binding is gone. Do not reintroduce either. Decommission checklist: docs/db-egress-relay.md.
  • sandbox/ - Agent skills, project scaffold templates (create-worker/), and the canonical validate-notebook.py (byte-copied into workers/main/analysis-sandbox-assets/ for the analysis image build context). Not the agent control plane or harness — those live in workers/main (chat-thread-do.ts, Pi tools, Dockerfiles).
  • scripts/ - Deploy, eval, self-host, and maintenance scripts.
  • docs/ - Supporting documentation; see docs/README.md for the canonical index (many *-plan.md / feedback files are historical).
  • plans/ - Active cross-cutting architecture plans (e.g. OrgDO split, no-VM build/deploy).
  • infra/ - Terraform for the static-IP database egress relay VM (infra/db-egress-relay/); infra/selfhost/ for self-host cloud templates. See infra/README.md.
  • tests/ - Vitest UI / src/lib unit tests (vitest.config.ts).
  • workers/main/tests/ - Worker / Durable Object / Miniflare tests + evals/ (vitest.workers.config.ts).
  • e2e/ - Playwright end-to-end specs.
  • .agents/skills/ - Agent skills for this repo (evals and shadcn).

Read the full file on GitHub · 397 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 Changed · +3 lines · +1,228 tokens per session efa4dd343b7d
  2. 5d ago First seen · 394 lines · 11,152 tokens per session scan A f9a09873dda1

Subscribe to this mod's changes

camelAI AGENTS.md is an instructions file published in the GitHub repository qaml-ai/camelAI (362 stars, last pushed 4d ago), licensed MIT. It adds 12,380 tokens to every session, about $0.0619 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 2026-08-30.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens