The Agent Skills Directory

[PROMPT_INJECTION]: The instructions focus on structured architectural analysis and do not contain patterns designed to bypass safety filters, impersonate administrative roles, or override system constraints.
[DATA_EXFILTRATION]: No commands for accessing sensitive local files (such as credentials, SSH keys, or environment files) or performing unauthorized network requests were found. The skill prioritizes internal documentation over external data transfer.
[OBFUSCATION]: A scan for Base64 encoding, hex sequences, zero-width characters, and homoglyphs yielded no results. All text is transparent and human-readable.
[REMOTE_CODE_EXECUTION]: The skill does not include any mechanisms for downloading or executing external scripts, nor does it define any third-party software dependencies.
[DYNAMIC_CONTEXT_INJECTION]: The skill does not employ the dynamic command execution syntax (!command) to run shell operations at load time.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to ingest and analyze technical context, it uses structured reasoning and formal documentation templates that do not present a vulnerable surface for malicious data interpolation.

architect