The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of security findings that could contain malicious instructions designed to influence the agent's behavior during the validation process.
Ingestion points: The agent ingests user-provided findings from pasted text or file paths as described in SKILL.md under the 'Input' and 'Step 1' sections.
Boundary markers: The instructions do not specify any delimiters or safety warnings to ensure the agent ignores instructions potentially embedded within the findings.
Capability inventory: The skill allows the agent to read source files (SKILL.md Step 2), perform network requests via the proxy skill and reaper CLI (SKILL.md Step 3), and append data to files (SKILL.md Step 6).
Sanitization: There are no documented sanitization or validation steps for the input finding data before it is used to guide the agent's analysis.
[COMMAND_EXECUTION]: The skill utilizes external command-line tools to perform live vulnerability testing.
Evidence: SKILL.md explicitly instructs the agent to execute reaper proxy to scope target domains and reaper get <id> to capture evidence of exploitability.

ghost-validate