The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands including ls, git log, and sqlite3 to aggregate project activity and token usage statistics for the review.
[DATA_EXFILTRATION]: The skill accesses personal files and sensitive paths, specifically Obsidian daily notes located in OneDrive and a local SQLite database (hannah.db) containing job metadata and costs.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data (git logs and personal notes) which provides a surface for indirect prompt injection if those sources contain malicious instructions.
Ingestion points: Git commit history from four local repositories and filenames/content from the user's daily notes directory (SKILL.md).
Boundary markers: No boundary markers or delimiters are used to isolate ingested data from the agent's instructions.
Capability inventory: The skill has permissions for shell command execution (ls, git, sqlite3) and the ability to write new files to the vault.
Sanitization: The skill does not implement any filtering or escaping of the content retrieved from external sources before presenting it to the agent.

weekly-review