Regex Patterns

**Judgment:** High-quality, broadly applicable regex cookbook that is very useful as a utility/reference skill for agents working with text, logs, or code. It’s strong on practicality and multi-language coverage, with good explanations of key regex concepts. **Strengths** - Rich set of ready-to-use patterns for the most common validation and parsing tasks. - Concrete examples for JavaScript, Python, Go, and CLI tools (grep/sed/Perl), making it easy for an agent to adapt patterns to different runtimes. - Includes guidance on flags, multiline behavior, escaping, greedy vs lazy, and backtracking/performance pitfalls. - Addresses common refactoring and text-cleanup workflows that agents often need when manipulating repositories or log files. **Risks / Limitations** - Some validation patterns are pragmatic rather than fully standards-compliant (e.g., emails, URLs, simplified IPv6); agents must not treat them as authoritative or complete RFC validators. - Regex-based validation and parsing can still produce false positives/negatives on edge cases; agents should be encouraged to pair them with higher-level logic when correctness is critical. - Performance and ReDoS risks exist if complex patterns are applied to large or untrusted input in tight loops, even though the skill does mention catastrophic backtracking and safer alternatives. - Regex is explicitly not suitable for full structural parsing of HTML/XML/JSON; agents must respect this and use proper parsers where appropriate. **Recommended Scenarios** - Agents doing everyday engineering support: log triage, quick data extraction from files, or one-off CLI text processing. - Codebase maintenance tasks: bulk renames, import path updates, or cleaning debugging statements using grep/sed/Perl. - Building lightweight input validation where perfect RFC adherence is not required (internal tools, admin UIs, scripts). - As a teaching/diagnostic aid when an agent needs to explain or fix a user’s regex or to avoid slow/unsafe patterns. Overall, this skill is a strong, general-purpose regex toolbox. It should be safe and effective if the agent is instructed to treat the patterns as pragmatic examples, not formal specifications, and to remain cautious about performance on large or adversarial inputs.