Borrowing it
Nothing to install: this file belongs to jordiboehme/crystalline. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/jordiboehme/crystalline/main/AGENTS.mdgit clone --depth 1 https://github.com/jordiboehme/crystallineWrote 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/jordiboehme/crystalline/agents-md)<a href="https://agentmods.dev/instructions/jordiboehme/crystalline/agents-md"><img src="https://agentmods.dev/badge/instructions/jordiboehme/crystalline/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/jordiboehme/crystalline/agents-md"><img src="https://agentmods.dev/badge/instructions/jordiboehme/crystalline/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.06659 | $0.06659 |
| Opus 5.5 | $0.02664 | $0.02664 |
| Sonnet 5.5 | $0.01332 | $0.01332 |
| Haiku 4.5 | $0.00666 | $0.00666 |
Grade A, and why
crystalline 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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Crystalline
Local-first knowledge management for humans and AI agents. Every Domain has exactly one source of truth: by default Markdown files (Engrams) on disk, or, for a virtual domain, the database itself. The search index is always a disposable derived layer; an MCP server and CLI sit on top.
Purpose
Crystalline gives an AI agent the capability to learn and evolve instead of starting from zero in every session. An agent is onboarded via a generated routing prompt, taught information through curated domains and stores its learnings and experiences as engrams while it works - becoming a useful and productive peer over time. All user-facing language (README, skills, MCP tool descriptions, routing prompt) is framed around onboarding, teaching, learning and experience rather than notes or documents.
Positioning follows the psychology concept of fluid vs crystallized intelligence: a language model is fluid intelligence (reasons in the moment, keeps nothing), Crystalline is the crystalline intelligence an agent accumulates and keeps, and the web UI (Fluid) is where the two meet. Describe the product as "crystalline intelligence"; the agent-facing phrase avoids the Crystalline/crystalline doubling and reads "Crystalline is your crystallized intelligence across sessions". Never describe it as "crystalline memory" or a memory tool.
Vocabulary
- Domain - a registered folder of knowledge with a mandatory MANIFEST.md at its root, used for routing
- Engram - one markdown file holding a unit of knowledge, with YAML frontmatter (OKF compatible)
- Address scheme:
crystalline://<domain>/<permalink>
Workspace
Cargo workspace, Rust edition 2024, pinned toolchain in rust-toolchain.toml.
crates/core(crystalline-core) - format layer: parser, emitter, schema, verify, prompt. Must never depend on async runtimes, databases or ML cratescrates/index(crystalline-index) - Store trait, embedded database backend, sync engine, search, embeddings. The local embedding model and the contradiction check's NLI model run on Metal on Apple Silicon Macs (candle'smetalfeature, enabled for aarch64-apple-darwin only) with a CPU fallback, and on the CPU everywhere else (src/device.rs, one device pick per model load)crates/remote(crystalline-remote) - GitHub-backed team collaboration plumbing and the client for connected Crystalline servers (credentials, sources, connection, cache): the forge-neutral Provider trait, the plain-text merge engine and the on-disk origin state. Git is never invoked; everything goes through the GitHub APIscrates/identity(crystalline-identity) - accounts, roles, domain access, joins and scopecrates/engine(crystalline-engine) - the shared engine, review-mode overlay, origins, co-editing sessionscrates/rest(crystalline-rest) - the JSON API, sign-in and the co-editing socketcrates/service(crystalline-service) - single-instance daemon, MCP server, control protocol; the facade the CLI and the integration tests reach everything throughcrates/cli(crystalline) - the single user-facing binary
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 Changed · +4 lines · +529 tokens per session 626772a55820
- 5d ago Changed · +5 lines · +1,071 tokens per session 612676ba3450
- 11d ago Changed · +2 lines · +165 tokens per session 7efd717a442e
- 13d ago Changed · +12 tokens per session a4d2766fb26b
- 14d ago Changed · +2 lines · +326 tokens per session 2f13f6a61b40
- 15d ago Changed · +1 lines · +208 tokens per session 7773deeff3bc
- 19d ago Changed · +1 lines · +44 tokens per session cc2e9bb4c1be
- 21d ago Changed · +16 tokens per session eec349a1c40a
- 22d ago First seen · 86 lines · 4,288 tokens per session scan A 8901ffd8a329
crystalline AGENTS.md is an instructions file published in the GitHub repository jordiboehme/crystalline (17 stars, last pushed 2d ago), licensed MPL-2.0. It adds 6,659 tokens to every session, about $0.0266 per session on Opus 5.5. 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 2026-10-06.
Other instructions, from other repositories
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).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
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 copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
vscode model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.