The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources which creates a potential surface for indirect instructions.
Ingestion points: The skill ingests data from external Frida documentation searches (search_frida_docs) and interprets output from instrumentation sessions as described in SKILL.md.
Boundary markers: The instructions do not define explicit delimiters or instructions to ignore commands potentially embedded within documentation or tool output.
Capability inventory: The skill possesses capabilities to write files and execute JavaScript code within target processes using the load_script tool.
Sanitization: There is no evidence of filtering or sanitization applied to the documentation results or process output before they are processed by the agent.
[DYNAMIC_EXECUTION]: The skill follows a pattern of script generation and subsequent execution.
In Phase 2: Scripting and Phase 3: Execution of SKILL.md, the agent is instructed to write JavaScript instrumentation scripts (based on templates/script-template.js) and execute them using a load_script tool.
This dynamic behavior is the primary purpose of the skill; it includes safety mitigations such as mandatory script_id declarations, idempotence guards to prevent duplicate hooks, and defined teardown paths (uninstall RPC export).

frida-mcp-workflow