sickn33/agentic-awesome-skills v18.11.0

v18.11.0 "Repository Engineering, Blueprint Planning, and a Cleaner Contributor Pipeline"

on GitHub

[18.11.0] - 2026-10-01 - "Repository Engineering, Blueprint Planning, and a Cleaner Contributor Pipeline"

Adds 8 reviewed skills - repository-native engineering, evidence-backed blueprint planning, and a five-skill English productivity pack - refreshes the spectral-field skill and the hosted catalog, and closes every open dependency advisory. Ships 2,610 skills.

A catalog release for Claude Code, Cursor, Codex CLI, Gemini CLI, and related AI coding assistants. It includes the complete protected maintainer batch merged after 18.10.0, with existing installation interfaces preserved.

Start here

• Install: npx agentic-awesome-skills@18.11.0
• Choose your tool
• Best skills by tool
• Bundles
• Workflows

Added

• Repository engineering pair (#1727) - repo-foundation implements features, fixes, and authorized contract migrations inside a repository's existing conventions, preserving user edits and establishing failure invariants before persisted-state changes. repo-native-refactor reviews a diff without editing, or performs evidence-based cleanup bounded by risk band R0-R4. Both are risk: critical and carry explicit stop conditions for unresolved ownership or contract consequences. MIT, adapted from Natchannnn/repository-engineering-skills.
• idea-to-blueprint (#1742) - turns a raw product idea into one evidence-backed Markdown build blueprint: intake questions, a tiered research protocol, ranked architecture drivers, a version-pinned stack, personas and flows, epics with Given/When/Then acceptance criteria, and a session protocol that lets a coding agent build one epic per fresh session. Ships a stdlib-only lint_blueprint.py that validates the finished document without executing it. risk: safe, MIT, authored by @iniesohidham.
• English productivity pack (#1741) - five focused English skills from alapha888/agent-skills-en (MIT): deep-research-framework tiers sources and cross-verifies before writing conclusion-first reports with explicit uncertainty; five-axis-code-review walks correctness, security, readability, performance, and test coverage and returns actionable comments instead of style nitpicks; git-commit-message writes conventional commits from the staged diff; meeting-notes turns raw notes into decisions, owned action items, and open questions; tech-writing-proofread returns an Original -> Suggestion -> Reason checklist without rewriting the document. All risk: safe.

Changed

• liuguang-banlan-ui spectral field (#1743) - refreshes the two parameterized UI modes with a calibrated palette contract, an sRGB-encoded shader with output-code dithering, a CSS fallback that stays hue-ordered instead of using fixed radial spots, continuous field time across pause/resume, and a repaint on resize. Adds a data-only manifest parser and validator, a deterministic measurement helper reporting OKLab deviation, lightness shift, and gray ratio, a caching-disabled preview server, and focused Python tests. risk: critical.
• Hosted catalog analytics (#1744) - installs @vercel/analytics and mounts <Analytics /> inside the router so page views are tracked across the catalog routes.
• youtube-summarizer output template (#1722) - switches the remaining Portuguese labels in the summary template to English, matching the rest of the template.
• React 19.3.0 (#1738, #1746) - aligns react and react-dom at 19.3.0 in the catalog web app and in the loki-mode example frontend, so both packages always share the same version.
• Top Contributor rankings (#1724) - adds npm run sync:top-contributors and its regression test, so the README ranking is recomputed from git history instead of being hand-maintained.
• Maintainer SkillSpector guidance (#1732) - calibrates how SkillSpector advisory findings are triaged in the canonical maintainer skill.

Fixed

• typescript-expert diagnostic script (#1730) - removes shell=True from ts_diagnostic.py, so TypeScript project paths are passed as arguments rather than interpolated into a shell command.

Documentation and community

• Added the Natchannnn/repository-engineering-skills, iniesohidham/idea-to-blueprint, and alapha888/agent-skills-en source credits to the README.
• Thanks to @iniesohidham for idea-to-blueprint, @Natchannnn for the repository-engineering pair, @alapha888 for the English productivity pack, @3516027002att-ui for the liuguang-banlan-ui spectral-field refresh, and @ManiShah7 for the youtube-summarizer template cleanup.
• Closed every open dependency-upgrade proposal as either merged or superseded, and the repository now reports zero open Dependabot alerts.

Validation scope

Exact-head maintainer review for changed skill content, protected source PRs, repository validation and tests, reference validation, docs security, protected CI and CodeQL, dependency review, canonical synchronization, release preflight, npm publication, and release provenance verification.