Recently updated open-source projects

agent-lsp orchestrates existing language servers and exposes their code intelligence to AI agents through MCP. It keeps language servers warm between sessions, combines many routine calls into batch operations, and supports workflows such as caller analysis and speculative edits. The project documents 65 tools, 24 agent workflows, and CI-verified support for 30 languages.

Flutter Markdown Editor opens existing .md files from the file explorer and creates new ones with formatting controls for headings, links, tables, and other Markdown elements. Users can switch between editor and preview or view both at once, with image preview, LaTeX rendering, themes, and a multilingual interface. The app is mainly designed for mobile use, though the repository also publishes builds for some desktop platforms.

Novel Deconstruct analyzes a web novel from a Fanqie link or pasted text and produces an 18-part craft report. It fetches and decodes chapters, scans sentence and dialogue metrics, breaks the story into scenes, and tracks structure, character arcs, value changes, and repetitive wording. Fanqie is the primary supported platform; access after the first ten chapters may require login or membership, and font decoding depends on OCR.

Describe the desired blog style and the skill creates a complete Gridea Pro theme, then runs scaffolding, syntax validation, and mock-data rendering before handoff. It supports Jinja2, Go Templates, and EJS, with output ready to copy into the `themes/` directory. The repository explicitly focuses on functional themes, so a separate frontend design skill is recommended for stronger visual direction.

b2b-pm-workbench helps product teams design and ship complex B2B software, especially SaaS and AI products with permissions, approval flows, and multi-tenancy. It combines more than 50 frameworks and 30 deliverables across discovery, solution design, delivery, growth, commercialization, and operations, with focused material on RBAC, ABAC, ReBAC, RAG, agents, pricing, and PLG.

iXBRL Skill helps accountants, auditors, reporting teams, and engineers navigate Inline XBRL requirements across major regulators and taxonomies. It combines primary-source references with Arelle workflows and a lightweight preflight script for common filing defects. Regulated submissions still require checking the current publisher rules and professional judgment.

Gitagent represents an AI agent as a Git repository, with model configuration, identity, rules, memory, tools, skills, hooks, and workflows stored as reviewable files. You can fork an agent, branch its personality, diff rule changes, and inspect memory history with normal Git operations. The runtime supports OpenAI, Anthropic, and OpenAI-compatible providers, and requires Node.js, Git, and the relevant API credentials.

py_ralph_frame provides a Python harness for spec-driven agent work: a plan defines the goal, specs define what to build, and each fresh-context iteration updates the implementation and project status. Pre-commit and pre-push gates can run linting, type checks, security scans, and tests for Claude, Codex, Gemini CLI, or Copilot; the harness intentionally does not create worktrees.

Use the corpus as a Claude Code skill, managed plugin, or MCP server in clients such as Cursor, Cline, Continue, and Zed. It guides work from elicitation and hierarchical backlogs through user stories, BDD, acceptance criteria, estimation, validation, and traceability, with Canadian English and Brazilian Portuguese editions. The project is intended for defining and checking requirements, not for implementing application code.

elevenlabs-v3-text-optimization converts flat text into ElevenLabs V3-ready scripts with emotional audio tags, pacing, punctuation, and contextual formatting. It respells names and brands for pronunciation, expands acronyms and numbers, removes unsupported SSML, and adapts the approach for dialogue, narration, technical, or dramatic delivery. Use it when the same words need clearer timing and more reliable synthesized speech.

Use this skill to audit and rewrite resume layout so systems such as Workday, Greenhouse, Lever, and iCIMS are more likely to extract company names, job titles, and dates correctly. It targets specific parsing failures such as year-only ranges, fused header lines, stylized company names, and parenthetical titles, then adds pre-upload and post-upload checks. It does not improve or invent resume content, and no format can guarantee a perfect parse.

VibePod gives you one CLI for launching Claude, Codex, Gemini, and other coding agents in separate Docker or Podman containers. It records usage and HTTP traffic locally, then exposes a dashboard for side-by-side comparisons. A compatible container runtime is required.

Give the skill a git repository and it turns the commit history into a single static kanban.html for PMs, QA, or business owners. The page groups 8–15 user-journey regression tasks, includes pass criteria and result capture, and keeps the full commit log with searchable impact tags and filters. It works offline without a build step, but it is an executable checklist rather than an automated test runner.

This repository is a working personal Claude Code configuration for Microsoft Fabric, Azure, Power BI, and related data-engineering tasks. You can cherry-pick individual skills, path-scoped language rules, hooks, a security-reviewer agent, MCP templates, and validation scripts instead of cloning the whole setup. It targets Windows, and every artifact is currently marked personal, with no semantic-versioning, compatibility, or support commitment.

VibeTree gives each AI-assisted task its own Git worktree so several features or fixes can progress in parallel with separate terminal state. Desktop, browser, and mobile access are available, but the main branch is still developing cloud and broader platform capabilities; the README recommends release-v0.1 for a stable desktop-only build.

Learn From Failure organizes sourced corporate case studies by failure mechanism, such as fraud, governance, cash burn, denial, or scaling too fast. Describe a decision and it can point to three to six relevant cases with the warning questions those companies raise, or you can browse the Markdown site and checklist without AI. It is a reflection tool, not a prediction engine, and its public cases should be checked against their sources.

lifeos-cli combines personal planning and evidence tracking in a local-first system for tasks, schedules, habits, notes, timelogs, relationships, and finance records. Its stable, help-driven command grammar works for both people and agents, while an optional local API and React interface operate on the same SQLite or PostgreSQL database and support portable exports, backups, and restores.

Upload a public-exam notice, choose the role or specialty, and generate the same structured syllabus as an Excel tracker, interactive HTML dashboard, and printable DOCX. The skill preserves the notice’s numbering and order rather than guessing which topics matter most. PDF extraction is heuristic, so the generated outline should still be checked.