The Agent Skills Directory

[SAFE]: The skill instructions emphasize a rigorous verification process for any data fetched from external sources, significantly reducing the risk of processing malicious or inaccurate information.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests untrusted data from the web using WebFetch and WebSearch (specifically from the sources listed in sources.md). However, this is mitigated by a mandatory evidence chain and verification discipline described in SKILL.md, which requires specific instrument numbers and primary source confirmation. This is consistent with the primary purpose of a research skill.
[COMMAND_EXECUTION]: The skill frontmatter allows for Read and Write tools, which are used to maintain a local baseline file (references/baseline.md). This file-writing behavior is restricted to the skill's own reference directory and is a documented part of the skill's intended update mechanism.

vietnam-crypto-radar