Public Agent Skills by anthropics on Lightlines.
Produces coherent, deduplicated answers from multi-source search results with attribution and confidence assessment.
Generates customized draft responses to common legal inquiries and identifies matters that require escalation.
Solves competition mathematics problems, adversarially verifies proofs, reports calibrated confidence, and optionally produces a verified PDF.
Integrates Model Context Protocol servers into Claude Code plugins to expose external service capabilities as tools.
Decodes workplace shorthand into context so the agent can understand and act on internal requests.
Guides agents through reading, creating, modifying, extracting, securing, and OCR-processing PDF files.
Creates self-contained interactive HTML playgrounds with visual controls, live previews, and copyable prompts.
Creates organized Claude Code plugins with standard layouts, manifests, components, paths, and discovery rules.
Creates a reusable skill from a user-provided PowerPoint template.
Creates, edits, reads, and validates PowerPoint presentation files.
Builds and publishes a private, shareable, tabbed status page for a multi-workstream project.
Reconciles account balances against supporting data and categorizes the differences found.
Generates an explorable, self-contained HTML report of Claude Code session usage in the current working directory.
Creates, improves, evaluates, benchmarks, and optimizes reusable skills.
Create effective skills for Claude Code plugins with structured content, progressive disclosure, and validation guidance.
Designs systems and evaluates architectural decisions through structured documents with assumptions and trade-off analysis.
Tracks, summarizes, adds, and completes tasks in a shared TASKS.md file.
Produces a prioritized technical-debt audit with effort estimates, business justifications, and a phased remediation plan.
Creates a structured feature specification or PRD from a problem statement or feature idea.