The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local scripts to execute system utilities for project and session management.
Evidence: scripts/new-worklog.ts executes git rev-parse --show-toplevel to locate the repository root.
Evidence: scripts/find-current-session.ts and scripts/get-session-transcript.ts invoke sqlite3 and opencode to search and export session history.
[EXTERNAL_DOWNLOADS]: The skill's scripts rely on the tsx package, which may be dynamically downloaded during execution.
Evidence: The scripts contain a shebang line #!/usr/bin/env -S npx tsx, which instructs the system to use npx to run tsx, potentially fetching it from the NPM registry if not present locally.
[DATA_EXFILTRATION]: The skill reads sensitive session history from the user's home directory to generate logs.
Evidence: scripts/get-session-transcript.ts accesses directories such as ~/.claude/projects and ~/.local/share/opencode to retrieve full session transcripts. Although the data is used locally, it involves reading the complete history of the user's interaction with the AI.
[PROMPT_INJECTION]: The skill processes session transcripts, which are untrusted data sources that could contain malicious instructions intended for the agent.
Ingestion points: Session transcripts are read from disk by get-session-transcript.ts.
Boundary markers: A sub-agent brief defined in assets/extraction-brief.md provides structural instructions and section headers for the extraction process.
Capability inventory: The skill has the capability to write the extracted content to the project's file system in the .worklogs/ directory.
Sanitization: No explicit sanitization or filtering of the transcript content is performed before it is provided to the sub-agent for processing.

worklog