The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses sensitive local files to function, including ~/.wakatime.cfg (which contains a plaintext API key), Claude Code history files, Codex session logs, and Cursor IDE SQLite databases. These files contain private conversation history and authentication credentials. The skill instructions also direct the agent to verify the WakaTime API key using cat, which exposes the secret directly to the agent's context.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Reads content from git commit messages and AI tool chat histories (Claude, Codex, Cursor) across multiple scripts.
Boundary markers: None identified; untrusted data is processed directly as strings.
Capability inventory: The skill can execute shell commands via git, make network requests to the WakaTime API, and write HTML/Markdown report files to the local filesystem.
Sanitization: The ingested text data from logs and commit history is not explicitly sanitized before being printed or embedded in reports.
[COMMAND_EXECUTION]: The script scripts/git_sessions.py uses subprocess.run to execute git log commands to extract session data. While it uses a list for arguments, it still represents a shell interaction surface. The skill also uses find to auto-discover git repositories in subdirectories.
[EXTERNAL_DOWNLOADS]: The scripts/wakatime_fetch.py script makes network requests to api.wakatime.com to fetch user summaries and durations. This is a well-known service associated with the skill's primary purpose.

time-lens