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/jitsucom/jitsu/agents-mdgit clone --depth 1 https://github.com/jitsucom/jitsuWhat 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 | $0.01596 | $0.01596 |
| Opus 5 | $0.00798 | $0.00798 |
| Sonnet 5 | $0.00319 | $0.00319 |
| Haiku 4.5 | $0.00160 | $0.00160 |
Grade A, and why
jitsu 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 — 176 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Jitsu — Agent Guidelines
Project Overview
Jitsu is an open-source data pipeline platform (self-hosted Segment alternative). It collects events from websites/apps and routes them to data warehouses and other destinations.
Repository Structure
This is a monorepo with two main technology stacks:
Go (/bulker)
Data ingestion engine for streaming events to warehouses.
bulkerapp/— main Bulker serviceingest/— HTTP ingest endpointsync-controller/— connector sync orchestrationbulkerlib/— core ingestion libraryconnectors/— warehouse connectors (ClickHouse, BigQuery, Redshift, Snowflake, S3, GCS, etc.)
Node.js / TypeScript
Services (/services)
rotor/— event routing, transformation, and function execution
Web apps (/webapps)
console/— admin UI (Next.js)
The enterprise-edition billing/admin API now lives in
jitsucom/jitsu-cloud-billing(deployed atbilling.jitsu.com).
Libraries (/libs)
jitsu-js/— browser JS SDK (@jitsu/js)jitsu-react/— React bindings (@jitsu/jitsu-react)functions/— functions runtime (@jitsu/functions-lib)juava/— shared utilities
CLI (/cli)
jitsu-cli/— developer CLI (jitsu-clion npm)
Types (/types)
protocols/— shared TypeScript protocols (@jitsu/protocols)
Tooling
- Node.js: pnpm ≥10 (workspace manager), Turbo (build orchestration), Node.js ≥22
- Go: Go 1.26 with Go workspaces (
go.workat repo root) - Frontend: Next.js, React 18, TypeScript, Tailwind CSS
- Testing: Jest (unit), Playwright (E2E), Go's built-in
testing - CI: GitHub Actions (
.github/workflows/)
Common Commands
# Install JS dependencies
pnpm install
# Generate Prisma client + zod schemas (required once after a fresh checkout
# or worktree). Skipping this leaves Turbopack panicking in
# ModuleGraphImportTracer::get_traces because it can't render the missing-
# module error for `prisma/schema`.
pnpm codegen
# Build all JS packages
pnpm build:turbo
# Type-check
pnpm typecheck:turbo
# Run tests. Console tests include an integration project that needs Docker
# (testcontainers: Postgres + ClickHouse); run only the pure tests with
# `pnpm exec vitest run --project unit` from webapps/console.
# Details: CONTRIBUTING.md "Testing".
pnpm test
# Lint / format
pnpm lint
pnpm format
# Start all dev services (hot-reload)
pnpm dev
# Start only the console
pnpm console:dev
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 · 176 lines · 1,596 tokens per session scan A e4b66bdcbd88
jitsu AGENTS.md is an instructions file published in the GitHub repository jitsucom/jitsu (5,057 stars, last pushed 3d ago), licensed MIT. It adds 1,596 tokens to every session, about $0.0080 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
mcp-toolbox GEMINI.md
Gemini CLI instructions for googleapis/mcp-toolbox, covering mcp toolbox context & style guide, project overview, tech stack, key directories and development workflow.
mcp-toolbox AGENTS.md
AGENTS.md instructions for googleapis/mcp-toolbox, a project described as: MCP Toolbox for Databases is an open source MCP server for databases.
clickhouse-go AGENTS.md
Instructions for ClickHouse/clickhouse-go, a project described as: Golang driver for ClickHouse.
mcp-toolbox CLAUDE.md
Claude Code instructions for googleapis/mcp-toolbox, a project described as: MCP Toolbox for Databases is an open source MCP server for databases.
orionbelt-semantic-layer AGENTS.md
AGENTS.md instructions for ralforion/orionbelt-semantic-layer, covering agents.md, project overview, commands, docker: two separate images (api + ui) and code review.
orionbelt-semantic-layer CLAUDE.md
Claude Code instructions for ralforion/orionbelt-semantic-layer, a project described as: Source-available semantic layer and sidecar for agentic AI and governed analytics. Compiles declarative YAML models into optimized SQL, KPIs, and semantic context across 8 SQL dialects.