Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-02-24 · source-backed
Domain: agent-security | Difficulty: intermediate | Source
Classify every skill: L0 (safe, read-only) → L3 (critical, shell/root/financial). 46.3% of skills are duplicates. Top 1% consume 100K+ tokens. Run uvx mcp-scan@latest --skills regularly. 76 confirmed malicious skills found across 3,984 analyzed. Sandbox all L2+ skills in production.
Each link below shares sources, entities, or timing with this story.
Shared entities / Shared topic / What happened next / Tension
Both cover Difficulty, Domain; overlapping topics (agent, agent-security, chain, difficulty); picks up the Difficulty thread on 2026-02-26.
Shared entities / Shared topic
Both cover Difficulty, Domain; overlapping topics (agent-security, critical, difficulty, latest, skill).
Shared entities / Shared topic / What happened next / Tension
Both cover Difficulty, Domain; overlapping topics (agent-security, difficulty); picks up the Difficulty thread on 2026-02-26.
Vercel released Sandbox / Shared entity: Sandbox / Shared topic / What happened next
Linked by a graph relationship (Vercel released Sandbox); both cover Sandbox; overlapping topics (agent, classification).
Shared entities / Shared topic / What happened next
Both cover Difficulty, Domain; overlapping topics (agent-security, intermediate, skill); picks up the Difficulty thread on 2026-03-01.
Both cover Difficulty, Domain; overlapping topics (agent-security, difficulty, intermediate); picks up the Difficulty thread on 2026-02-26.
Both cover Difficulty, Domain; overlapping topics (difficulty, intermediate, skill); picks up the Difficulty thread on 2026-02-26.
Both cover Difficulty, Domain; overlapping topics (agent, classification, critical); picks up the Difficulty thread on 2026-02-26.