scriptorium
AGENTS.md1 starsChanged 59 days agoMITView on GitHub
Fail-closed guardrails for agent-written Obsidian vaults. A scriptorium was where scribes copied manuscripts under strict rules, with correctors verifying the work before it left the room. The repository ships three things: a PreToolUse hook that refuses four classes of vault-corrupting write before they reach disk, the laws those checks enforce, and a leak scanner that keeps personal material out of published rule text. Read LAWS.md before writing into a vault. It is the complete rule set and the specification the…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,542 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository5 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- vault-conventionsSkill
- vault-logSkill
- vault-upSkill
- vault-wrapSkill
The file
We split this AGENTS.md into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
Contributing: never commit personal material.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Scriptorium
Fail-closed guardrails for agent-written Obsidian vaults. A scriptorium was where scribes copied manuscripts under strict rules, with correctors verifying the work before it left the room.
The repository ships three things: a PreToolUse hook that refuses four classes of vault-corrupting write before they reach disk, the laws those checks enforce, and a leak scanner that keeps personal material out of published rule text.
Read LAWS.md before writing into a vault. It is the complete rule set and the specification the hook is written against. The laws are not summarized here — a summary of a law is the failure mode this project exists to prevent.
Laws are parameterized by .scriptorium.json at the vault root; copy .scriptorium.example.json and edit. The vault root is wherever that file lives, so there is no path to configure. Outside a directory containing one, the hook does nothing and the laws do not apply.
Four further laws are documented separately: AI-Write Boundaries, Provenance and the Truth Hierarchy in AUTHORSHIP.md; Vault Structure in the shape spec named by shape_spec.
Contributing: never commit personal material. scripts/leak-scan.py runs as a pre-commit hook (git config core.hooksPath .githooks) and rejects the commit on any hit.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.