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 skills add jordiboehme/crystalline --skill crystalline-intelligencegit 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/skills/jordiboehme/crystalline/crystalline-intelligence)<a href="https://agentmods.dev/skills/jordiboehme/crystalline/crystalline-intelligence"><img src="https://agentmods.dev/badge/skills/jordiboehme/crystalline/crystalline-intelligence/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/skills/jordiboehme/crystalline/crystalline-intelligence"><img src="https://agentmods.dev/badge/skills/jordiboehme/crystalline/crystalline-intelligence.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 | $0.00036 | $0.06650 |
| Opus 5.5 | $0.00014 | $0.02660 |
| Sonnet 5.5 | $0.00007 | $0.01330 |
| Haiku 4.5 | $0.00004 | $0.00665 |
Grade A, and why
crystalline-intelligence 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 5d 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Crystalline Intelligence
Crystalline is your crystallized intelligence across sessions: what you have been taught and what you have learned lives in Domains as Engrams you search, read and, where allowed, capture into.
Start from the routing block
At connection you are handed a routing block as your instructions: one line per registered domain plus the tool names it routes through. Treat each line as a targeting aid, not a full catalog - an unrelated-looking domain may still hold the answer. Re-fetch the same index mid-session with list_domains and include_routing: true.
On a shared instance that block may be a count rather than a list, or may name fewer domains than exist: an instance can hold private domains, and what you are shown is what the account you connected as may see. So list_domains with include_routing: true is the index, not a refresh of one you already have - call it at session start whenever the instructions carried no domain lines.
Domains you may not see
A domain the account you connected as is not a member of is not there for you at all: it is absent from list_domains and from every search, and naming it directly answers exactly as naming a domain nobody ever registered. Read a "not registered" answer as "not yours to see or not there, and you cannot tell which" - never as a cue to create it with add_domain, and never as a bug to report. Ask the person you are working with for access instead.
The same rule governs ending a domain. remove_domain unregisters one - the counterpart to add_domain, and what to call when someone says a domain is finished rather than telling them to open a terminal - and on your own machine you may remove any of them. On a shared instance it takes an instance admin or, for a private domain, its owner; anything else is refused with text naming who can, which is an answer to relay rather than a call to retry. It never deletes a file domain's markdown, and it refuses a virtual domain (whose engrams are only in the database) until you pass purge: true, so confirm the loss with the user first. On a harness that can ask, the call also puts one question to the person, with its checkbox already checked: Accept removes the domain and Decline keeps everything.
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.
- 5d ago Changed 44304fb7b640
- 8d ago Changed 084fc51321e7
- 13d ago Changed · +2 lines ac9fb3b6222f
- 14d ago Changed · +2 lines 4b14be94582d
- 15d ago Changed af655b3c5544
- 16d ago Changed · +2 lines 2a2b298cc9d9
- 17d ago Changed · +4 lines b6fbe3bd758b
- 18d ago Changed dad79bbd38c5
- 19d ago Changed a1a5c528cf02
- 21d ago First seen · 73 lines · 36 tokens per session scan A 030aec01751c
crystalline-intelligence is a skill published in the GitHub repository jordiboehme/crystalline (17 stars, last pushed 2d ago), licensed MPL-2.0. It adds 36 tokens to every session and 6,650 once invoked, about $0.0001 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 2026-08-31.
Other skills, from other repositories
jumbo-add-guideline
Use liberally when the user expresses a preference about how work should be done — coding style, process, testing approach, communication style. Captures the guideline so future sessions follow it without being told again.
jumbo-add-invariant
Use liberally when you discover or are told a non-negotiable constraint — something that must always be true regardless of context. Registers the invariant so future sessions never violate it.
jumbo-context-maintenance
Use when implementation discovers durable project context that must be recorded in Jumbo. Captures decisions, components, dependencies, relations, guidelines, and invariants during the work.
jumbo-correction-capture
Use when the user corrects an approach, constraint, workflow, or style. Converts corrections into Jumbo guidelines or invariants so future sessions receive the constraint.
atomicmemory
Persistent semantic memory across Claude Code sessions — user preferences, project context, prior decisions, codebase facts. Call memorysearch before answering questions that reference past work. Call memoryingest after the user shares durable facts.
atomicmemory-cli
Use the installed AtomicMemory CLI for memory search, ingestion, packaging, diagnostics, and agent-safe JSON output.