The Agent Skills Directory

[METADATA_POISONING]: There is a discrepancy between the reported author (qinyan-ai) and the author listed in the SKILL.md metadata (K-Dense Inc.).
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface because it processes untrusted external data (such as molecular Hamiltonians and circuit parameters) that could contain malicious instructions designed to influence the agent's behavior during code generation or simulation execution.
Ingestion points: Data enters the context via problem definitions described in references/algorithms.md and references/patterns.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded instructions within processed quantum data.
Capability inventory: The skill enables the agent to generate and execute Python code for quantum simulations and hardware interactions.
Sanitization: There is no evidence of input validation or sanitization for external parameters before they are used in code execution blocks.
[EXTERNAL_DOWNLOADS]: The skill recommends installing standard, well-known packages from the Qiskit ecosystem, including qiskit, qiskit-ibm-runtime, and qiskit-aer. It also mentions a non-standard "Samplomatic" package, although no installation command for it is provided.

qiskit