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/joelborch/flow/claude-mdgit clone --depth 1 https://github.com/joelborch/flowWhat 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 | $0.00736 | $0.00736 |
| Opus 5 | $0.00368 | $0.00368 |
| Sonnet 5 | $0.00147 | $0.00147 |
| Haiku 4.5 | $0.00074 | $0.00074 |
Grade A, and why
flow CLAUDE.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 — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Flow
A fast, minimal, agent-first task manager (self-hosted ClickUp/Asana
replacement). Everything runs on Cloudflare: one Worker (apps/api) serving
the SPA, REST, MCP at /mcp, and WebSockets at /ws; one Durable Object
(packages/core) holding the entire workspace in SQLite; R2 for attachments;
Queues for outbound side effects. Full agent contract: AGENTS.md.
Ground rules
packages/sharedis the contract. All entity shapes, mutation inputs, delta events, WS protocol,LIMITS, and MCP tool names live there as Zod schemas with inferred types. Import from it; never redeclare its types. If the contract must change, make the smallest compatible change and flag it prominently.- One workspace, one DO instance. The DO is the only writer; REST/MCP
handlers call its RPC methods (
WorkspaceRpc) and never touch SQL. - Every mutation appends a Delta with a monotonic seq; WS clients receive patches and replay on reconnect. Never build a "refetch the board" path.
- Automations evaluate inline in the DO mutation turn (depth cap 5). Outbound
side effects (webhook/email) go through the
SIDE_EFFECTSqueue, never inline. Email respectsEMAIL_DRY_RUN(default true): log, don't send. - Subtasks are Asana-style: done/not-done + optional assignee/dueDate. They do NOT have statuses.
- API keys (
flow_bearer tokens) impersonate a real user; the audit trail records the key id. - TypeScript strict everywhere. pnpm workspaces. No new runtime deps without a strong reason (allowed: hono, zod, preact, @preact/signals, the MCP SDK, or similarly tiny libraries).
- UI: clean, minimal, Linear-ish. Tailwind v4. Fast: optimistic mutations, CSS transforms for drag, no full-board re-renders.
- Never commit secrets or real config values; placeholders only
(
flow.example.com,[email protected]). Local values go inapps/api/.dev.vars.
Layout
packages/shared— the Zod contract (depends only on zod)packages/core— Workspace Durable Object + automation engine; most testsapps/api— Hono Worker: auth, routes, MCP, WS, queue consumer, wrangler.jsoncapps/web— Preact + @preact/signals + Tailwind v4 SPA; store insrc/storeapps/importer— local-only ClickUp ETL (tsx, never deployed)
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 · 55 lines · 736 tokens per session scan A d294ee33abb1
flow CLAUDE.md is an instructions file published in the GitHub repository joelborch/flow (0 stars, last pushed yesterday), licensed MIT. It adds 736 tokens to every session, about $0.0037 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.
Other instructions, from other repositories
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.
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).
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.
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.
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).
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.