Brief
v1.0.1Condense information into actionable briefings. User specifies sources, skill structures the output.
⭐ 3· 1.1k·9 current·9 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description match the declared behavior: structuring user-provided information into briefs and learning format preferences. The skill requires no credentials or external binaries and its actions (formatting + storing preferences) are reasonable for the stated purpose.
Instruction Scope
SKILL.md limits data access to user-provided sources and explicitly says it will not access files/email/calendar without user request — scope is appropriately constrained. One ambiguity: the mechanism by which the agent will obtain user-granted access to sources is not specified, so ensure the agent prompts and does not assume broad permissions.
Install Mechanism
Instruction-only skill with no install spec or downloaded code — lowest-risk delivery model. All content is present in the provided SKILL.md and supplemental docs.
Credentials
No environment variables, credentials, or external config paths are requested. The only local resource used is a directory under the user's home (~ /brief) for storing preferences and templates, which is proportional to its stated behavior.
Persistence & Privilege
The skill persists format preferences to ~/brief/preferences.md and templates under ~/brief/templates. This is a limited, local persistence model and not a broad system-wide privilege; users should be aware these files will be created/updated in their home directory.
Assessment
This skill is instruction-only and appears to do what it says: format user-provided input into briefs and keep simple, local preference data under ~/brief. Before installing or using it, consider: (1) you're consenting to the skill creating/updating ~/brief/preferences.md and ~/brief/templates — back up any existing folder with that name if needed; (2) only grant access to concrete sources when asked (avoid blanket folder/account permissions); (3) because the skill source/homepage are unknown, review the SKILL.md and templates yourself and confirm you’re comfortable with the stated behavior; (4) if you want stricter safety, require that the agent prompt you before accessing any file or external service and avoid granting long-lived credentials or global file access.Like a lobster shell, security has layers — review code before you run it.
latestvk975krtxf9c97f40svmcay47w1819r4y
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
📋 Clawdis
OSLinux · macOS · Windows