The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests untrusted data from project files.
Ingestion points: The skill reads primary sources from docs/task/{ID}-{task-name}.md and docs/testing/{ID}-{task-name}.md to determine documentation content.
Boundary markers: No boundary markers or instructions to ignore embedded commands within the ingested files are present.
Capability inventory: The skill has the capability to write to critical project files like CLAUDE.md, LEARNINGS.md, and TASKS.md, as well as creating new documentation files.
Sanitization: There is no evidence of sanitization or validation of the content read from the task or testing documents.
[COMMAND_EXECUTION]: The skill executes shell commands to inspect the state of the repository.
Evidence: Uses git diff --name-only main...HEAD to identify file changes and cross-reference them against the task document.

document