The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local wrapper script (scripts/burp-search.sh) which invokes Burp Suite (Java) with user-provided arguments to query project data.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external Burp Suite project files (.burp), which may contain attacker-controlled HTTP traffic payloads.
Ingestion points: Untrusted data enters the agent context via the output of burp-search.sh when querying proxyHistory or siteMap components in SKILL.md.
Boundary markers: The skill instructions recommend truncation using head and jq, but there are no explicit boundary delimiters or instructions to ignore embedded commands within the processed traffic data.
Capability inventory: The skill possesses the Bash and Read tools, enabling it to execute local shell commands and read files.
Sanitization: The burp-search.sh script implements a robust awk filter that restricts standard output to valid JSON objects only. Additionally, it truncates diagnostic/error output to a maximum of 20 lines and 500 characters per line, providing a defense-in-depth mechanism against large payloads or malicious data leaking into the agent's context through logs.

burpsuite-project-parser