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/kozakhou/project-bourne/agents-mdgit clone --depth 1 https://github.com/KozakHou/project-bourneWrote 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/kozakhou/project-bourne/agents-md)<a href="https://agentmods.dev/instructions/kozakhou/project-bourne/agents-md"><img src="https://agentmods.dev/badge/instructions/kozakhou/project-bourne/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 | $0.03694 | $0.03694 |
| Opus 5 | $0.01847 | $0.01847 |
| Sonnet 5 | $0.00739 | $0.00739 |
| Haiku 4.5 | $0.00369 | $0.00369 |
Grade A, and why
project-bourne 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 — 907 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project Bourne — Agent Instructions
Project Identity
Project: Project Bourne
Python distribution: bourneprov
Python package: bourneprov
CLI: bourne
Tagline:
Every experiment has a history.
Short description:
Universal experiment provenance and reproducibility for science and engineering.
Project Bourne is a general-purpose scientific experiment provenance, reproducibility, traceability, and durable record-keeping system.
Its purpose is to preserve how scientific and engineering results came to exist.
The core abstraction is the experiment, not the programming language, framework, solver, or AI agent.
Core Product Principle
Universal by default, domain-aware when available.
Bourne core must work without knowing whether an experiment uses:
- Python
- JAX
- PyTorch
- TensorFlow
- Julia
- MATLAB
- R
- C
- C++
- CUDA
- Fortran
- MPI
- COMSOL
- OpenFOAM
- LAMMPS
- GROMACS
- proprietary scientific software
- shell scripts
- arbitrary executables
Framework- and domain-specific knowledge belongs in optional collectors, adapters, integrations, or later verification modules.
Mental Model
Git answers:
How did this code come to exist?
Bourne answers:
How did this scientific result come to exist?
A result may depend on:
- source code
- Git state
- data
- configuration
- parameters
- environment
- dependencies
- compiler/runtime
- hardware
- operating system
- execution command
- random state
- upstream experiments
- generated artefacts
- human actions
- agent actions
Bourne exists to preserve that history.
Non-Negotiable Rules
-
Never make Bourne core dependent on a specific ML framework.
-
Arbitrary commands and executables are first-class experiments.
-
Basic provenance capture must not require users to modify their scientific source code.
-
A failed experiment is still an experiment and must be recorded.
-
Provenance correctness takes priority over UI polish.
-
Bourne must remain useful without AI agents.
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 · 907 lines · 3,694 tokens per session scan A 55409d8910f4
project-bourne AGENTS.md is an instructions file published in the GitHub repository KozakHou/project-bourne (0 stars, last pushed 6d ago), licensed Apache-2.0. It adds 3,694 tokens to every session, about $0.0185 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
clio-kit CLAUDE.md
Claude Code instructions for iowarp/clio-kit, covering claude.md, project overview, project structure, common development commands and setup & installation.
srunx CLAUDE.md
Instructions for ksterx/srunx, covering claude.md, development commands, package management, cli usage and testing.
SurfaceFluxes.jl AGENTS.md
AGENTS.md instructions for CliMA/SurfaceFluxes.jl, covering surfacefluxes.jl agent guide, ecosystem guidelines, before you act: agent autonomy, repo-specific guidelines and architecture.
VibeCodeHPC CLAUDE.md
Instructions for Katagiri-Hoshino-Lab/VibeCodeHPC, covering vibecodehpc, project overview, language priority, design principles and package structure.
bio-gene-to-reference-tree copilot-instructions.md
Copilot instructions for Hongda-Zhao/bio-gene-to-reference-tree, a project described as: Auditable agent skill for resolving protein queries, selecting references, and planning reproducible phylogenetic trees.
braina GEMINI.md
Instructions for brainets/braina, covering project: braina (brain interaction analysis), 1. project context & purpose, 2. commands, verify environment (all core dependencies) and run the verification test suite for frites + hoi.