The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local shell script (scripts/whittle.sh) to perform static analysis and rendering of Markdown-based decision trees. The script is self-contained and uses awk for parsing, avoiding dangerous functions or external command execution.\n- [EXTERNAL_DOWNLOADS]: The skill documentation includes a link to the author's GitHub repository (github.com/0froq/skills-pack) for source information. No remote code or assets are downloaded or executed by the skill during runtime.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes repository files (e.g., decisions.md) to extract design questions and decisions. \n
Ingestion points: Markdown files specified by the user as arguments to the script.\n
Boundary markers: Content is structured using a specific grammar (headings for IDs, prefix labels for fields like status: and edges:).\n
Capability inventory: The agent executes scripts/whittle.sh which reads the file and performs basic parsing and logic checks via awk.\n
Sanitization: The Bash script employs an awk parser that specifically looks for defined field markers and sanitizes input (e.g., removing tabs from titles) before processing.\n- [SAFE]: No malicious patterns, such as obfuscation, credential exfiltration, or unauthorized persistence, were identified. The skill's behavior is consistent with its stated purpose of managing design documentation.

whittle