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/robofinsystems/robosystems/claude-mdgit clone --depth 1 https://github.com/RoboFinSystems/robosystemsWrote 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/robofinsystems/robosystems/claude-md)<a href="https://agentmods.dev/instructions/robofinsystems/robosystems/claude-md"><img src="https://agentmods.dev/badge/instructions/robofinsystems/robosystems/claude-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 | $0.06081 | $0.06081 |
| Opus 5 | $0.03040 | $0.03040 |
| Sonnet 5 | $0.01216 | $0.01216 |
| Haiku 4.5 | $0.00608 | $0.00608 |
Grade A, and why
robosystems CLAUDE.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 today.
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.
- **Frontend-facing auth** (JWT/Bearer) is a frontend concern; backend testing with `curl` should use the API key. How it starts
The opening of the file, as written. The whole thing — 546 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md - RoboSystems Development Guide
Critical Rules
All Python commands MUST use uv run:
uv run python script.py # NOT: python script.py
uv run pytest # NOT: pytest
uv run ruff check # NOT: ruff check
Always use Docker profile robosystems - never individual service profiles:
just start # Uses robosystems profile by default
just start robosystems # Explicit form (same result)
# NOT: just start api
Never use os.getenv() directly - use centralized config:
from robosystems.config import env
database_url = env.DATABASE_URL # NOT: os.getenv("DATABASE_URL")
Never create migrations manually - always autogenerate:
just migrate-create "description" # NOT: manual alembic revision
API Endpoints (local testing)
Core platform APIs are mounted under /v1 (graphs, billing, auth, etc.).
Extensions (roboledger, roboinvestor) live under /extensions and are
graph-scoped at the URL level, with three sub-surfaces:
- Typed reads →
POST /extensions/{graph_id}/graphql(Strawberry + GraphiQL in dev) - Command writes →
POST /extensions/{roboledger|roboinvestor}/{graph_id}/operations/{operation_name} - Analytical view operations →
POST /extensions/{domain}/{graph_id}/operations/{view_name}— graph-backed operations that query LadybugDB rather than the extensions OLTP database. Read-only, same envelope contract as command writes.build-fact-gridis the first one (pivot tables over the XBRL hypercube); gated independently of the OLTP domain flags so deployments without the corresponding tenants can still mount them
All three surfaces take graph_id as a URL path parameter — auth + per-graph
access are validated by FastAPI dependencies before the handler runs.
GraphQL queries do NOT take a graphId argument; the URL is the scope.
Per-domain feature flags: ROBOLEDGER_ENABLED and ROBOINVESTOR_ENABLED
gate the corresponding GraphQL resolvers and operation routers. The
schema is built dynamically per flag combo, so a ledger-only deployment
exposes only ledger fields (no INVESTOR_NOT_INITIALIZED runtime errors).
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.
- today Changed · +2 tokens per session 540eba1b67db
- 4d ago First seen · 546 lines · 6,079 tokens per session scan A 4e2cb4cdecff
robosystems CLAUDE.md is an instructions file published in the GitHub repository RoboFinSystems/robosystems (24 stars, last pushed yesterday), licensed Apache-2.0. It adds 6,081 tokens to every session, about $0.0304 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.
Other instructions, from other repositories
James-RAG-Evol CLAUDE.md
Claude Code instructions for Hashevolution/James-RAG-Evol, covering sekos / james — session briefing for claude code, what james is, where we are right now, critical rules for this session and where to look next.
auto_trader CLAUDE.md
Instructions for mgh3326/auto_trader, covering claude.md, 프로젝트 개요, 개발 환경 설정, 필수 요구사항 and 초기 설정.
auto_trader AGENTS.md
Instructions for mgh3326/auto_trader, covering agents.md — 외부 에이전트(codex 등) 진입점, 하드룰 — 절대 위반 금지, 최소 명령어, 더 읽을 것 and 유지 규약.
WorldOfTaxonomy CLAUDE.md
Instructions for colaberry/WorldOfTaxonomy, covering claude.md - worldoftaxonomy, what this project is, architecture, tech stack and how to run.
another-s3-manager CLAUDE.md
Instructions for kruchenburger/another-s3-manager, covering another s3 manager, stack, structure, database and migration commands.
loopback-training-server CLAUDE.md
Instructions for aderaaij/loopback-training-server, covering training api, tech stack, project structure, frontend (web dashboard) and development.