All categories
AI & Agents
LLMs, agent workflows, RAG, MCP, prompting, and AI app patterns
- 304,408
- 12,684
Security grades appear on each card once the skill has been scanned. Newly imported skills may briefly show without a grade until the backfill job runs.
Open in full browserBrowse ai & agents skills
Showing 1,609–1,632 of 304,408 skills
- 140 Claude F42a86f3**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude Ea7360b1AI model provider implementations organized by capability type.Votes: 0GitHub stars: 9
- 140 Claude D93fd1c2**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude C78ef70egit fetch origin main git log --oneline origin/main -3 # See if your PR was merged git checkout main git pull origin main git checkout -b claude/new-feature-<session-id> ``` - Claude Code web sessions can span user interactions where PRs get merged between messages - The user may merge and delete branches without explicitly telling Claude - Attempting to push to a deleted branch will fail with 403 errors - Always verify branch state before assuming continuityVotes: 0GitHub stars: 9
- 140 Claude C3e94863**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude C01f7c6agit fetch origin main git log --oneline origin/main -3 # See if your PR was merged git checkout main git pull origin main git checkout -b claude/new-feature-<session-id> ``` - Claude Code web sessions can span user interactions where PRs get merged between messages - The user may merge and delete branches without explicitly telling Claude - Attempting to push to a deleted branch will fail with 403 errors - Always verify branch state before assuming continuityVotes: 0GitHub stars: 9
- 140 Claude Bbc8ba8f**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude 9a830ca9**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude 8d4c1b0bgit fetch origin main git log --oneline origin/main -3 # See if your PR was merged git checkout main git pull origin main git checkout -b claude/new-feature-<session-id> ``` - Claude Code web sessions can span user interactions where PRs get merged between messages - The user may merge and delete branches without explicitly telling Claude - Attempting to push to a deleted branch will fail with 403 errors - Always verify branch state before assuming continuityVotes: 0GitHub stars: 9
- 140 Claude 803f2e10**VERSION:** See SKILL.md frontmatter ---Votes: 0GitHub stars: 9
- 140 Claude 702150daImmutable domain model. No database access, no HTTP, no external dependencies. Downstream layers (services, infrastructure) consume these types.Votes: 0GitHub stars: 9
- 140 Claude 6e623341Long-term memory system. Batch extraction from conversation segments, entity linking, consolidation, and proactive retrieval. All services created and managed by `LTMemoryFactory`.Votes: 0GitHub stars: 9
- 140 Claude 6dbc39daThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.Votes: 0GitHub stars: 9
- 1178 Name Skill B4462054Scaffold repo-local Ralph autonomous harness under scripts/ralph/. Use when user runs /flow-next:ralph-init.Votes: 0GitHub stars: 9
- 1178 Name Skill 15916ba5This skill should be used when the user asks to "repair an agent", "audit an agent", "fix my agent", "review agent quality", "check if my agent is well-written", "diagnose agent problems", "what's wrong with this agent", "improve this agent", or "what's wrong with this agent file".Votes: 0GitHub stars: 9
- 1178 Name Skill 07dab008Inverted orchestration—Codex CLI acts as CTO (planning and reviewing), Claude Code executes with native tools. Use when a task benefits from external architectural oversight before and after implementation.Votes: 0GitHub stars: 9
- 1153 Workflow 30cfd097Follow these steps in order. This workflow is **idempotent** - safe to re-run.Votes: 0GitHub stars: 9
- 1153 Workflow 058cd28eFollow these steps in order. This workflow is **idempotent** - safe to re-run.Votes: 0GitHub stars: 9
- 1097 Troubleshooting 50903b7f> This guide covers common issues when using EvalView for AI agent testing, including type errors, connection problems, framework-specific issues, and evaluation debugging.Votes: 0GitHub stars: 9
- 1097 Troubleshooting 271bcac1This guide covers common issues when using EvalView and how to resolve them.Votes: 0GitHub stars: 9
- 1087 Tools 84fc5444OpenBotX is an AI assistant platform. Tools are Python functions the AI agent can call during conversations. This document provides a complete reference for all built-in tools, their parameters, and access rules.Votes: 0GitHub stars: 9
- 1082 Todo 9674497f- [x] **Fix YAML frontmatter parsing** — Removed HTML comments before `---` in 8 files (from @kylewhirl fork) - [x] **SSRF prevention in Python scripts** — Private IP blocking in fetch_page.py and analyze_visual.py (from @artyomsv #7) - [x] **Install hardening** — venv-based pip, no `--break-system-packages` (from @JawandS #2) - [x] **Windows install fixes** — `python -m pip`, `py -3` fallback, requirements.txt persistence (from @kfrancis #5, PR #6) - [x] **requirements.txt persistence** — CopieVotes: 0GitHub stars: 9
- 1050 System Prompt 0ec11ccb始终仔细思考。保持耐心和彻底。不要过早放弃。 始终，保持极其简单。不要过度复杂化事情。 从零开始完成任务时，你应该： - 理解用户的需求 - 制定实现计划 - 逐一利用已有工具和能力完成任务 在处理现有代码库时，你应该： - 理解代码库和用户的需求。确定最终目标以及实现该目标的最重要标准。 - 对于错误修复，通常需要检查错误日志或失败的测试，扫描代码库以找到根本原因，并找出修复方法。如果用户提到任何失败的测试，你应该确保更改后这些测试能够通过。 - 对于功能开发，通常需要设计架构，并以模块化和可维护的方式编写代码，同时最小化对现有代码的侵入。如果项目已有测试，请添加新测试。 - 对于代码重构，如果接口发生更改，通常需要更新所有调用你正在重构的代码的地方。不要更改任何现有逻辑（尤其是在测试中），只专注于修复由接口更改引起的任何错误。 - 为实现目标进行最少的更改。这对你的性能非常重要。 - 遵循项目中现有代码的编码风格。 你的回答应该简洁明了，直奔主题。 除非用户要求详细解释，否则你必须用不超过 4 行的文字（不包括工具使用或代码生成）简洁地回答问题。 重要提示：在保证答案的实用性...Votes: 0GitHub stars: 9
- 055 Architecture Cbe3b86c``` ┌──────────────────────────────────────────────────────────┐ │ vLLM API Layer │ │ (OpenAI-compatible: chat, completions, embeddings, │ │ audio, tools, MCP, reasoning) │ └──────────────────────────────────────────────────────────┘ │ ▼ ┌──────────────────────────────────────────────────────────┐ │ MLXPlatform │ │ (Votes: 0GitHub stars: 9