The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the bash tool. In Step 0, it executes ${QMD_BIN} update && ${QMD_BIN} embed, where QMD_BIN is a variable dynamically read from a local configuration file (.claude/bulletjournal.local.md). Additionally, Step 5 explicitly instructs the agent to run qmd get via shell to retrieve project status.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it aggregates content from various markdown files into a structured review.
Ingestion points: Reads multiple files from daily/ and 10-Projects/ directories (SKILL.md steps 3, 4, and 5).
Boundary markers: None. The skill does not instruct the agent to treat the content of the notes as data only or to ignore embedded instructions.
Capability inventory: File system write access (Write tool used in step 8) and shell command execution (bash tool used for qmd commands).
Sanitization: None. The skill does not perform any validation, escaping, or filtering on the content retrieved from the daily notes before processing it or including it in the final summary output.

weekly-review