data-security
Skill7 starsChanged 19 days agoMITView on GitHub
Check whether data is safe to share with an external tool. Say "is this safe to paste", "can I send this to <tool>", or "classify this data". Fires before pasting, uploading, forwarding, or sending data to an external tool or AI model. Classifies the data and checks it against the user's data-handling policy.
In this repository100 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- notion-package/CLAUDE.mdCLAUDE.md
- llms.txtllms.txt
- add-handsSkill
- add-mouthSkill
- add-voiceSkill
- approval-gatesSkill
- auditSkill
- backupSkill
- blind-spot-reviewSkill
- bottleneck-diagnosticSkill
- brain-logSkill
- brain-passSkill
- brain-snapshotSkill
- brand-interviewSkill
- brand-voice-interviewSkill
- business-context-loaderSkill
- campaign-from-themeSkill
- capture-sweepSkill
- catch-upSkill
- client-updateSkill
- connectSkill
- content-repurposerSkill
- content-startSkill
- context-persistenceSkill
- cross-linkSkill
- decision-frameworkSkill
- decisions-startSkill
- delivery-startSkill
- email-drafterSkill
- employee-reviewSkill
- finance-importSkill
- forcing-questionsSkill
- founder-coachingSkill
- founder-next-moveSkill
- founder-os-setupSkill
- founder-reviewSkill
- founder-scope-challengeSkill
- github-opsSkill
- handoff-protocolSkill
- hireSkill
- historySkill
- housekeepingSkill
- ingestSkill
- knowledge-captureSkill
- legal-complianceSkill
- linkedin-brand-directionSkill
- linkedin-network-scanSkill
- linkedin-postSkill
- linkedin-power-auditSkill
- linkedin-startSkill
- linkedin-warm-revivalSkill
- lintSkill
- list-prunerSkill
- log-archiveSkill
- log-replySkill
- meeting-prepSkill
- memory-passSkill
- menuSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 26
Read the first available policy source.
- Distinctive: line 33
If none of the above exist or yield a policy, use the default matrix below and tell the user it is a fallback.
- Distinctive: line 45
If ambiguous, choose the more restrictive class.
- Distinctive: line 78
Do not silently redact and proceed.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
Data Security
Runs on: reasoning - reads your files and reasons; any capable agent can run this.
Block unsafe data movement before it leaves the user's control.
When To Run
- The founder is about to paste content into an AI tool.
- A file is about to be uploaded to an external service.
- Client, employee, financial, legal, health, or personal data is involved.
- The tool is new, unlisted, or consumer-grade.
- The user asks "can I paste this" or "is this safe to upload".
Pre-Read
Read the first available policy source. Skip a source silently if it does not exist; do not error out.
rules/data-handling.md(likely absent on a fresh install - skip if missing)rules/approval-gates.md(or fall back totemplates/rules/approval-gates.md)stack.json(lists which tools the user has approved)context/clients.md(per-client restrictions, if any are noted)
If none of the above exist or yield a policy, use the default matrix below and tell the user it is a fallback.
Data Classes
| Class | Examples | Default Tool Access |
| Public | Published posts, public pages, public case studies | Any tool |
| Internal | SOPs, internal notes, non-sensitive business plans | Paid tools with business terms |
| Client general | Client name, project scope, public deal existence | Tools with no-training terms or business agreement |
| Client sensitive | Strategy, financials, private messages, personal data | Approved tools only |
| Regulated | Health, legal, banking, tax, ID, payroll, protected personal data | Specific approved systems only |
If ambiguous, choose the more restrictive class.
Tool Classes
| Tool Class | Examples | Allowed Data |
| Approved secure | Enterprise or API tools listed in the user's policy | Up to the policy limit |
| Business paid | Paid tools with no-training or business terms | Public, internal, client general if policy allows |
| Consumer or free | Free chat tools, unapproved browser tools | Public only |
| Unknown | Anything not listed | Public only |
Procedure
- Identify the data involved.
- Classify the highest-risk data class.
- Identify the destination tool.
- Classify the tool class from policy or fallback.
- Compare data class to tool class.
- Return
ALLOW,REDACT FIRST, orBLOCK. - For
BLOCK, name a safer tool or a local-only path.
Output Format
Data class: <class>
Tool class: <class>
Call: ALLOW | REDACT FIRST | BLOCK
Reason: <one sentence>
Safe path: <tool or local action>Redaction Rule
Do not silently redact and proceed. Show what type of information must be removed, then ask the user to confirm the redacted version before using it.
Rules
- Client-sensitive and regulated data never go to unknown tools.
- Per-client restrictions override the generic matrix.
- A one-time exception needs explicit human approval and a log entry.
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.