The Agent Skills Directory

[Prompt Injection] (LOW): Indirect Prompt Injection Surface. The skill ingests untrusted user descriptions to generate functional Cairo smart contract code.
Ingestion points: User-provided inputs in 'Interactive' and 'Direct' modes defined in SKILL.md.
Boundary markers: Absent. The skill does not instruct the agent to use delimiters or ignore instructions embedded within user descriptions.
Capability inventory: The skill uses Write and Edit tools to persist generated logic to the filesystem.
Sanitization: Absent. There is no input validation or sanitization of user-provided natural language before it is used to generate code.
[Data Exfiltration] (SAFE): No sensitive file access or network communication patterns detected. No hardcoded credentials were found.
[Remote Code Execution] (SAFE): No remote code downloads or execution of untrusted scripts identified.
[Obfuscation] (SAFE): No evidence of Base64, zero-width characters, or other obfuscation techniques found in the skill content.
[Privilege Escalation] (SAFE): No commands related to privilege escalation (e.g., sudo, chmod) were detected.

dojo-system