The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bash script in SKILL.md using the pattern bash "${CLAUDE_PLUGIN_ROOT}/scripts/camel-springboot-matrix.sh" ${ARGUMENTS:-4.14.0}. The use of ${ARGUMENTS} without double quotes allows for shell word splitting and glob expansion. If a user provides input containing shell metacharacters (e.g., ;, &&, |, or backticks), it can lead to arbitrary command execution on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an ingestion surface for untrusted data by downloading and parsing HTML and XML files from external repositories.
Ingestion points: The script scripts/camel-springboot-matrix.sh fetches release data and POM files from https://repo.maven.apache.org and https://camel.apache.org.
Boundary markers: The skill does not use specific delimiters or instructions to prevent the agent from following potential instructions embedded in the generated compatibility matrix.
Capability inventory: The skill has the capability to perform network requests (curl) and write files to the local directory (mkdir, echo).
Sanitization: The skill uses grep and tr to extract XML property values. While this provides basic text extraction, it does not fully sanitize the data against complex injection patterns in the downloaded XML content.
[EXTERNAL_DOWNLOADS]: The script scripts/camel-springboot-matrix.sh performs multiple network operations to fetch metadata and POM files from Apache's Maven repository and the official Camel website. These references to well-known services are documented as follows:
Fetches Camel release metadata from https://camel.apache.org/releases/.
Downloads project object model (POM) files and directory listings from https://repo.maven.apache.org/maven2/.

camel-matrix