The Agent Skills Directory

[SAFE]: The skill does not contain any malicious code, obfuscation, or unauthorized data access patterns. Its primary function is to automate the detection of project metadata from standard manifest files.
[PROMPT_INJECTION]: No malicious prompt injection patterns were detected. The skill includes explicit defensive constraints in SKILL.md (e.g., "NO silent writes", "ALWAYS require explicit user approval", and the use of a pre-tool-use-validator.sh hook) to prevent unauthorized modifications to the codebase. It utilizes a mechanical detection contract defined in references/detection.md to ensure that stack claims are evidence-based rather than generated via model inference.
[DATA_EXFILTRATION]: The skill reads local project manifest and configuration files (e.g., package.json, pyproject.toml, go.mod) as part of its core functionality to identify the technology stack. This data is used solely to generate a local configuration proposal at scratchpad/tailor-proposal.md and is not transmitted to external domains.
[COMMAND_EXECUTION]: The skill defines procedures for scanning files and generating text but does not include instructions to execute arbitrary shell commands or external scripts. All file operations are restricted to reading project configuration and writing to a designated scratchpad file.

tailor