The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches a list of TOR exit nodes from the official Tor Project domain (check.torproject.org). This operation is consistent with the skill's stated purpose of cross-referencing network connections against known TOR infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external network connection logs, creating a potential surface for indirect instructions to be processed by the agent.
Ingestion points: scripts/agent.py reads external log data through the parse_zeek_conn_log and parse_netflow_csv functions.
Boundary markers: The parsing logic does not implement specific delimiters or instructions to ignore potential commands embedded within the connection log fields.
Capability inventory: The script includes capabilities to write a JSON report to the local file system and perform network GET requests to fetch threat intelligence.
Sanitization: Log data is parsed and used for statistical calculations without explicit sanitization or filtering of string-based fields.
[METADATA_POISONING]: There is a discrepancy between the author 'mahipal' listed in the SKILL.md frontmatter and the author 'mukul975' identified in the LICENSE copyright notice.

analyzing-ransomware-network-indicators