The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external sources like PRDs, Slack threads, and meeting transcripts to extract ticket information. This creates an attack surface for indirect prompt injection where malicious instructions could be embedded in the processed text.
Ingestion points: Workflow B and Workflow C in SKILL.md ingest user-pasted text for ticket extraction.
Boundary markers: The skill includes references/UNTRUSTED_CONTENT.md, which explicitly instructs the agent to treat such content as data rather than instructions and to avoid merging it with the agent's own output.
Capability inventory: The skill uses the cawplan CLI (via the Bash tool) to list products and create tickets.
Sanitization: The skill mandates a "Confirm before creating" step for all ticket creation actions and includes detailed rules to ignore any instructions found within the data.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute cawplan CLI commands. These commands are used for legitimate workflows such as authenticating, searching for products/versions, and creating tickets as requested by the user.

cawplan-ticket-create