Hybrid scanner finds malicious AI agent skills with fewer false positives
The full pipeline detects 68 of 80 malicious skills and cuts false positives from 40 to five versus Cisco's LLM mode.
Stories per day · 30 days
2 RSS
Relatedsupply-chain4clawhub2llm-agents2ai-agents1cisco-skill-scanner1compositional-risk1cosa1gitskills1information-flow1jailbreak1knotscryer1langchain1language-agents1malware-detection1markdown1mcode1minimax1open-source1
19 stories, newest first · one card per story
Stories tagged #agent-skills
MiniMax open-sources MCode Stack of 15 agent skills
The supplied page text is a cookie notice and does not describe the skills or license.
Scanner finds local and composite vulnerabilities in public agent skills
Analysts confirmed 53 of 59 GitHub findings, and mitigations blocked attacks in 47 of 55 tested cases.
Markdown skill files propagate jailbreaks into AI agents
A poisoned SKILL.md raised mean attack success from 0.5% to 45% across seven open-weight models.
About 88,000 agent skills contain oversight-loosening directives
Roughly one in seventeen GitSkills files tells agents to skip checks, and a few upstream families supply half the copies.
CoSA study shows composed agent skills raise policy-violation rates
Skill additions capped at 5% of each pool raised violations from 35.02% to 58.45% across five models.
Skill-infinity makes agent skill optimization self-improving
On 83 SkillsBench tasks the authors report 60.7% pass rate, above curated skills and SkillEvolver.
SCS finds 627 malicious agent skills in 66,157 ClawHub packages
On a balanced 234-skill benchmark it reached 85.6% accuracy, 95.2% recall and 86.9% F1.
KnotScryer finds 612 risky chains among clean agent skills
On 13,598 ClawHub skills, 13.7% of flagged chains were made only of skills a single-skill scanner labeled safe.
Socratic Skill Learning interleaves teaching with agent practice
Fixed-Window SSL scored 0.801 utility on 155 cases, above no-Skill, Trace2Skill, and SKILL-KD baselines.
Most agent-skill listings carry undeclared elevated capabilities
Of 6,309 marketplace listings, seven in ten had an elevated capability, and 59% of declarations understated the bundle.
SkillMarker: A KGW-Based Watermarking Method for Agent Skill Ownership Verification
SkillMarker watermarks agent skills with KGW so owners can verify copies without breaking procedural accuracy.
Skill Constellations: Tracing the Supply Chain of Agent Skills on GitHub
Study of 2.2 million GitHub agent-skill copies finds stars miss the repositories that spread risk.
Runaway Reaction: When Benign Skills Compose into Malicious Behavior
Paper finds vetted benign agent skills can compose into malicious behavior and introduces a 4,000-skill benchmark.
SkillSeek: Revisiting Agent Skill Retrieval at Marketplace Scale
SkillSeek shows standard retrieval matches LLM-mediated agent skill selection at about half the cost.
SKILLLITE: Evidence-Guided Malicious Skill Auditing with Compact LLMs
SKILLLITE audits third-party Agent Skills for hidden malicious behavior using compact local LLMs.
3
Placeholder third-party[.]com Referenced Across 1,700+ Repositories Now Serves Malicious Content
Squatted placeholder domain third-party.com now serves a Windows ClickFix lure referenced in over 1,700 repositories.
4
Cyberthreats are moving faster than SMBs: Readiness must accelerate
ESET's SMB Cyber Readiness Index 2026 finds 73% of SMBs adopting AI while 40% lack AI policy, and reports 3,000+ malicious AI agent skills.
1
Ask HN: How do you manage skills files?
A Hacker News thread debates whether agent skill files are worth managing, citing 2–4x output-token reductions on flagship models in one company's testing.
4