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/jonashertner/opencaselaw/agents-mdgit clone --depth 1 https://github.com/jonashertner/opencaselawWrote 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/jonashertner/opencaselaw/agents-md)<a href="https://agentmods.dev/instructions/jonashertner/opencaselaw/agents-md"><img src="https://agentmods.dev/badge/instructions/jonashertner/opencaselaw/agents-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 | $0.00537 | $0.00537 |
| Opus 5 | $0.00269 | $0.00269 |
| Sonnet 5 | $0.00107 | $0.00107 |
| Haiku 4.5 | $0.00054 | $0.00054 |
Grade A, and why
opencaselaw 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 4d ago.
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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md - OpenCaseLaw Codex Guidance
OpenCaseLaw is a production Swiss legal corpus and public API. Treat this repo as live legal infrastructure: favor evidence, small reversible changes, and explicit escalation over broad autonomous edits.
Bootstrap
Before non-trivial maintenance work, read:
CLAUDE.md~/.claude/projects/-Users-jonashertner-caselaw-repo-1/memory/MEMORY.md- Relevant memory topic files from that index
TECHNICAL_OVERVIEW.txtdocs/agent-loop/LOG.md
Verify memory claims against current code, logs, or production probes before using them as ground truth.
Hard Rules
- Production read DBs must be opened
mode=ro&immutable=1. - Never add write paths to live
decisions.db,reference_graph.db, ordecision_structure.db. - Preserve atomic-swap rebuild behavior; do not replace it with in-place DB writes or naive copies.
- All FTS5 user input must pass through
_sanitize_fts5. - Do not construct legal citation strings or direct quotations yourself. Use the project citation and verbatim-text sources.
- New scraper data paths must use official sources, not entscheidsuche. Retire
es_*feeds only after verified direct parity and no coverage loss. - Tests stay offline unless a target explicitly says it is a live check.
- Never read or print secrets from
.env*, crontab, systemd environment files, or production config.
Autonomous Boundary
Autonomous Codex work may measure, document, improve read-only monitoring, add offline tests, and make small non-gated fixes after verification.
In autonomous mode, stop at a proposal for changes touching:
publish.py,build_fts5.py, DB schemas,base_scraper.py- new scrapers or material scraper behavior changes
- QC blocking behavior
state/**, production output DBs, billing, secrets, backup destinations, systemd deployment behavior, or paid resources
Use ops/autonomy-policy.json and scripts/agent_safe_deploy.py as the
mechanical deployment guard. No commit, push, or deploy without explicit user
approval plus passing make test and make verify-offline.
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.
- 4d ago First seen · 60 lines · 537 tokens per session scan A 134f57c02ef6
opencaselaw AGENTS.md is an instructions file published in the GitHub repository jonashertner/opencaselaw (60 stars, last pushed today), licensed MIT. It adds 537 tokens to every session, about $0.0027 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
ClawBench AGENTS.md
AGENTS.md instructions for TIGER-AI-Lab/ClawBench, covering clawbench -- agent context, what this is, project structure, setup and 2. configure at least one model.
agentcounsel AGENTS.md
AGENTS.md instructions for zgbrenner/agentcounsel, covering agents.md — agentcounsel, what this repository is, first, decide which job you are doing, job 1 — using a skill and non-negotiable rules.
OpenCUA CLAUDE.md
Instructions for xlang-ai/OpenCUA, covering claude.md, project overview, commands, environment setup and model inference dependencies.
Legal-Workflow-Chain CLAUDE.md
Instructions for lawyerllsx-dotcom/Legal-Workflow-Chain, covering claude.md — llm wiki schema（开源模板）, 核心原则, 文件处理规则, 文书完成前自检清单 and 知识存储规则.
fedlex-mcp CLAUDE.md
Instructions for malkreide/fedlex-mcp, covering claude.md, teil 1 — portfolio-konventionen, vor der arbeit, tests and wenn etwas rot ist.
Legal-Agent-Skills CLAUDE.md
Instructions for code-lawyer/Legal-Agent-Skills, covering claude.md, 本仓库是什么, 先读：作者宪章（principles.md）, 常用命令 and 结构校验——改动任一 skill 的 references/ 或 rules/ 后必跑，绿了才提交.