The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided ABAP code to perform reviews, which presents a surface for indirect prompt injection attacks.
Ingestion points: The workflow in SKILL.md involves reading ABAP code provided by the user for both automated linting and manual Clean ABAP review.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the code content being processed.
Capability inventory: The skill possesses the ability to execute shell commands (npm, abaplint) and access the file system within the project root to read code and write configuration files.
Sanitization: There are no explicit sanitization or filtering steps for the input code before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill manages the installation of and references to external tools and libraries.
Evidence: The skill instructions (SKILL.md and references/abaplint.md) suggest installing the @abaplint/cli package via NPM.
Evidence: The configuration reference (references/abaplint.md) points to external GitHub repositories for ABAP type definitions, such as github.com/abaplint/deps and github.com/abapedia/steampunk-2302-api. These are recognized community resources in the ABAP ecosystem.
[COMMAND_EXECUTION]: The skill uses local shell commands to perform static analysis and environment setup.
Evidence: It executes commands like abaplint, npx @abaplint/cli --version, and npm install @abaplint/cli -g to manage and run the linting process.

abap