The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to download an external Python script from a remote repository not associated with a trusted organization.
Evidence: curl -O https://raw.githubusercontent.com/k-kolomeitsev/data-structure-protocol/main/skills/data-structure-protocol/scripts/dsp-cli.py in SKILL.md.
[REMOTE_CODE_EXECUTION]: The skill executes the downloaded script locally, which could allow arbitrary code execution if the remote source is compromised or malicious.
Evidence: python dsp-cli.py --root <project-root> <command> in SKILL.md.
[PROMPT_INJECTION]: The skill processes untrusted codebase data to build structural memory, creating a surface for indirect prompt injection attacks.
Ingestion points: The dsp-cli.py script traverses the entire project directory, reading source code, configuration files, and assets to generate UIDs and descriptions (SKILL.md).
Boundary markers: Absent. The instructions do not specify any delimiters or warnings to prevent the agent from following instructions embedded within the codebase entities being mapped.
Capability inventory: The skill possesses file system read and write capabilities, and the ability to execute the dsp-cli.py script via the command line (SKILL.md).
Sanitization: Absent. There is no mention of filtering, validation, or sanitization of the content extracted from the project files before it is stored in the .dsp/ directory or processed by the agent.

data-structure-protocol