moru-python
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill recommen d s the in st al l at io n of the
morupac k ag e via pip. Sin c emoruis no t fr om a t r us t ed or g an iz at io n lis t ed in the sec ur it y g uid el in es, it is fl ag g ed as an un v er if iabl e dep en d en c y. How ev er, as it is the p r imar y p ur p os e of the skill, the sev er it y is red uc e d.- [COMMAND_EXECUTION] (SAFE): The skill fac il it at es remo t e comman d execu t io n wit hin a clou d san d bo x (sb x.comman d s.run). This is exp ec t ed beh av io r fo r the des c r ib e d us e cas e.- [PROMPT_INJECTION] (LOW): The skill p r es en t s a v ul n er abil it y sur f ac e fo r In d ir ec t P r omp t In j ec t io n (Cat eg or y 8). \n - In g es t io n p oin t s: The skill rea d s un t r us t ed dat a fr om the san d bo x via
sb x.fil es.rea d(),vol.dow n l oa d(), an d comman d s t d ou t s t r ea m s. \n - Bou n d ar y mar k er s: No del imit er s or war n in g in st r uc t io n s ar e p r ov id e d to sep ar at e p r oc es s e d dat a fr om ag en t in st r uc t io n s. \n
- Capabil it y in v en t or y: The ag en t can exec ut e ar bit r ar y comman d s, wr it e fil es, an d man ag e en v ir on men t var iabl e s wit hin the san d bo x. \n
- San it iz at io n: No san it iz at io n of ext er n al con t en t is p r es en t in the examp l e s.- [SAFE] (SAFE): The aut omat ed sc an rep or t reg ar d in g
vol.deap p ear s to be a fal s e po s it iv e cau s e d by a reg e x mis p ar s in g the Py t ho n met ho dvol.del et e()inSK I L L.md.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata