The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using the bash environment:
A curl POST request to http://localhost:31337/notify for voice notifications.
Execution of a TypeScript linter located at ~/.claude/LIFEOS/TOOLS/SkillDriftLint.ts via bun.
Appending log data to ~/.claude/LIFEOS/MEMORY/SKILLS/execution.jsonl using echo.
[DATA_EXFILTRATION]: The skill performs local network operations by sending JSON payloads to localhost:31337. While targeted at the local machine, this is a network request triggered during skill execution.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes untrusted instruction files from the filesystem.
Ingestion points: settings.json, CLAUDE.md, and various instruction files discovered at runtime.
Boundary markers: Absent; there are no specific delimiters or instructions for the agent to ignore embedded commands within the files being audited.
Capability inventory: The skill possesses file read/write capabilities and shell command execution (curl, bun, bash).
Sanitization: Absent; the skill does not appear to sanitize or validate the content of the files it evaluates before processing them.

BitterPillEngineering