The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to interact with issue trackers via bd and gh (GitHub CLI) commands. It constructs shell commands by interpolating strings extracted from the requirement document (PRD) directly into command arguments.
Evidence: The instructions in SKILL.md specify: bd create "PRD-FR-001: <requirement title>" -d "$(cat body.md)" -p 2. If a requirement title contains shell metacharacters (e.g., backticks, semicolons, or command substitutions), they could be executed by the system shell.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (the PRD markdown file) which can influence the agent's actions.
Ingestion points: The skill uses the Read tool to ingest the entire content of a file provided at <path to PRD.md>.
Boundary markers: No specific boundary markers or "ignore instructions" delimiters are defined for the ingested file content.
Capability inventory: The skill has access to Bash (shell execution), Read (file access), and Grep. It can create issues in external trackers (GitHub and Beads).
Sanitization: The instructions explicitly mandate copying requirement text and acceptance criteria "verbatim" and "unchanged," which prevents the model from filtering potentially malicious payload instructions or escape characters embedded in the PRD.

prd-to-issues