The Agent Skills Directory

[SAFE]: The skill's primary purpose is to define a persona focused on technical honesty and evidence-based assessment. All analyzed files consist of benign instructions and testing scenarios.
[NO_CODE]: The skill package does not contain any executable scripts, binaries, or package manifests. It is composed strictly of documentation, evaluation configuration, and plain-text references.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided claims and technical suggestions (Ingestion points). While no technical boundary markers or specific sanitization procedures are defined, the skill leverages investigation tools for code, docs, and diagnostics (Capability inventory) to verify inputs. This 'Trust but verify' approach, which mandates checking objective sources rather than blindly following user assertions, serves as a functional mitigation against indirect prompt injection by encouraging skepticism and independent verification.

professional-honesty