The Agent Skills Directory

[DATA_EXFILTRATION]: The skill contains absolute file paths to a local user directory in the 'References' section (e.g., /Users/khanm/...). This reveals the author's local system structure and username, which is a form of metadata exposure.
[COMMAND_EXECUTION]: The skill executes shell-based commands for syntax checking and building code, including python -c, npx tsc, cargo check, and dotnet build. While intended for quality assurance, these commands operate on files potentially modified or created by other skills or external inputs.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted data from multiple sources to guide its execution loop.
Ingestion points: Reads data from JIRA ticket descriptions, Confluence specifications, and output from other engineer skills.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands within the processed data are defined.
Capability inventory: The skill has the ability to execute shell commands (compilers/linters), perform git operations (commit/diff), and interact with JIRA (transitions/logging).
Sanitization: There is no evidence of sanitization or validation of the content retrieved from external JIRA/Confluence sources before it is used to formulate prompts for delegation or review.

sprint-executor