Loading...
Loading...
Found 2,167 Skills
Feishu/Lark CLI - official open-source CLI tool from Feishu for AI Agents. Provides 200+ commands across 12 business domains: IM, Docs, Sheets, Base (Bitable), Calendar, Video Meeting, Mail, Tasks, Wiki, Drive, Contacts, Search. Supports both user identity and bot identity authentication. Use when user wants to operate Feishu/Lark resources.
The social network for AI agents. Post, comment, upvote, and create communities.
Generate or update project memory for AI agents — default to AGENTS.md, support agent-specific targets such as CLAUDE.md, and keep sibling memory files synchronized while capturing stable architecture, conventions, and operational knowledge
Use when agents must debate, conference, deliberate, or reach consensus on a goal — competing positions argue and converge on one deliverable, adversarial review with synthesis, multi-stakeholder deliberation, structured disagreement with a forcing-field deliverable. Triggers: 'have agents debate X', 'reach consensus on Y', 'argue distinct positions and converge'. Not for saved team configs, agents/<slug> artifacts, implementation, or open-ended research.
Use this skill when users work within a local, single-user, compound-growth Markdown personal wiki (Karpathy's 'LLM owns wiki' model) — covering: ingesting raw/ materials (papers/ clippings/ external repo symlinks), querying and cross-page synthesis/ contradiction reconciliation, archiving conclusions back to the wiki, linting orphan/ outdated summaries, and format upgrades. Three core rules: raw/ controlled by users + wiki/ owned by LLM + AGENTS.md as the single source of truth. Triggers: "Ingest this paper into the wiki" / "Does the wiki have/ summarize content about X?" / "There's a contradiction between A and B in the wiki" / "Save the previous conclusion to the wiki" / "Check the wiki for orphan pages/ outdated summaries" / "Upgrade the wiki/ check wiki version" / "Include repo X in the wiki". Always use this skill whenever users need to digest materials/ retrieve wiki deposits/ archive new conclusions — even if they don't mention the skill name. Not applicable to: cloud/ team wikis (Notion/ Confluence/ Outline, etc.); wiki metadata configuration, wiki creation/deletion, session start/stop (use a single llmw command directly). **Trigger only when the cwd is the wiki root (containing `wiki_metadata.toml` + AGENTS.md skeleton)**; cross-wiki/ workspace operations go to `yzr-llm-workspace-management`; not applicable to other directories.
Build AI features with the first-party Laravel AI SDK (Laravel 13+); agents, embeddings, images, audio, and tool calling with provider-agnostic APIs
Guide for Infisical Privileged Access Manager (PAM) — brokering human and AI-agent access to databases, servers, Kubernetes clusters, and cloud accounts without the connecting party ever seeing a credential, with full session recording and audit. Covers all 13 account types (SSH, PostgreSQL, MySQL, MSSQL, OracleDB, MongoDB, Redis, Kubernetes, AWS IAM, GCP service account, Azure CLI, Windows, Windows AD), the accounts/folders/templates/memberships model, Admin/Connector/Auditor roles, session lifecycle and recording, just-in-time access requests with approvals, account credential rotation, discovery, dependencies, web and CLI access, and agentic access for AI agents via `infisical pam agentic access`. Use this skill when someone asks about: Infisical PAM, privileged access, session recording, just-in-time database access, brokered SSH access, giving an AI agent database access safely, access requests and approvals for infrastructure, or 'how do I let someone into production without giving them the password'. For humans and AI agents reaching infrastructure without ever holding a credential. Not for applications fetching a credential themselves (infisical-dynamic-secrets).
Map of AG2 capabilities and which sibling skill to reach for. Load first when the user mentions building with AG2 (ag2) but the specific feature isn't yet clear — agents, tools, model config, delegation, memory, observers, structured output, HITL, AG-UI, MCP server hosting, A2A protocol, realtime voice, telemetry, testing, or evaluation.
Single-agent recursion and parallel fan-out within one AG2 `Agent` — auto-injected `run_subtask` / `run_subtasks(parallel=True)` (opt in via `tasks=TaskConfig(...)`) for self-delegation, and `Agent.as_tool()` as a lightweight no-hub way to call one named agent from inside another. Use when a single coordinator wants to break work into its own sub-tasks, fan out concurrent sub-tasks, or invoke a specialist agent as a tool. Covers context flow, recursion safety, and `persistent_stream` for sub-task history. **For two or more agents actually collaborating with a registry, durable channels, governance, or turn-taking, use `ag2-network-quickstart` instead** — the network is the standard multi-agent pattern in AG2.
Build realtime voice / live audio agents with AG2's `ag2.live` module. Wrap a prompt + provider config in `LiveAgent` and open a bidirectional voice session with `agent.run()`, pumping mic audio in and playing synthesized speech out. Covers the two realtime providers — Gemini Live (`GeminiRealTimeConfig`) and OpenAI Realtime (`OpenAIRealTimeConfig`) with audio/text output modalities, voices, and user-speech transcription; audio I/O over the local sound card (`SoundDeviceRecorder` / `SoundDevicePlayer`, both sounddevice-backed); one-shot speech-to-text (`OpenAITranscriber`, `OpenAITranslationTranscriber`) and its `.pipe(agent)` voice pipeline; text-to-speech (`OpenAITTSConfig`); and `TTSObserver`, which speaks a regular text `Agent`'s streamed tokens aloud. Use when the user wants a talking agent, a phone/voice assistant, live transcription, or to add TTS playback to a text agent.
Compare AG2 agents, models, or prompts to decide which is better. run_variants scores several named agents on one suite and ranks them on a leaderboard (Variants holds a mapping of named Agent instances plus an axis label). run_pairwise with pairwise_judge does head-to-head LLM comparison using a dual-order position swap (a win counts only if it survives the swap, else a tie), reporting win-rate with a Wilson 95% CI, wins, losses, ties, flips, and agreement (Cohen's kappa). human_pairwise collects a person's blinded vote inline, or via an exported manifest with export_pairwise_cases and human_labels. Use when the user wants to A/B test prompts or models, run a leaderboard, pick a winner, judge head-to-head, measure win-rate, or collect human preference labels. For running and grading a single agent, see ag2-evaluation.
Test AG2 agents and tools without hitting a real LLM provider. Pass `TestConfig(...)` from `ag2.testing` as the agent's config (or per-`ask`) to mock LLM responses, inject `ToolCallEvent`s to simulate tool execution, and assert success / error paths. Use when the user is writing pytest tests for an Agent or Tool.