Unlazy Agent Skill

by Leonxlnx• 4.2k•• Scanned•Claim this listing•Request removal
Productivity & Workflowai agentsclaudeclaude codellmproductivity

Anti-laziness skill for AI agents. Core: the Depth Tree method, which splits a task N layers deep and gives every leaf the full time budget of the whole task, so effort multiplies with depth. Grounded in 2025-2026 research on model laziness, underthinking and premature completion.

Unlazy is an open-source agent skill by Leonxlnx, written mainly in JavaScript, with 4,227 GitHub stars and the MIT license. Its code was last updated Sep 3, 2026.

Scan summary: scanned Nov 5, 2026 at commit 1667149. Seen able to read environment and secrets. Scan policy

aboutSKILL.md / READMEsecurityreviews
Scanned
Full scan at 1667149 · Nov 5, 2026
Scanned Builds
Static code scanpass
Secrets scanpass
Dependency audit0 known vulnerabilitiespass
ManifestSKILL.mdpass
Buildpass
Boot & handshaken/a
Network egressn/a
Reproducible (×3)n/a

Findings (1)

  • low
    Reads environment variables not listed in the manifest
    PATHEXT, UNLAZY_SHELL, UNLAZY_APPROVAL_DIR, UNLAZY_SCOPE

Automated result for the commit shown. See our scan policy. View the scan log

Maintainers can claim the listing to rescan, or dispute a result.

Unlazy alternatives

View more →

Similar agent skills in Productivity & Workflow, ranked by GitHub stars. Compare licence, recent activity and scan status before you choose.

NameStarsLicenceLast updatedScan
Unlazy (this page)
Leonxlnx
4.2kMITSep 3, 2026Scanned, no flagged issues
Eigent
eigent-ai
16kApache-2.0Nov 1, 2026Scanned, no flagged issues
Kami
tw93
12kMITNov 1, 2026Scanned, no flagged issues
Skill
anbeime
9kNone foundNov 1, 2026Not scanned yet
Refly
refly-ai
7.5kNone foundOct 21, 2026Not scanned yet
N8n Skills
czlonkowski
6.6kMITOct 24, 2026Scanned, no flagged issues