Loading...
Loading...
Found 24 Skills
Orchestrates the full journey from zero to a running Neo4j application. Executes 8 named stages in order: prerequisites → context → provision → model → load → explore → query → build. Each stage has its own reference file in references/ that the agent reads and follows when entering that stage. Supports both HITL and fully autonomous operation. Time budget: ≤15 min after DB is running (autonomous), ≤90 min total (HITL).
Guides edge and tactical autonomous systems—perception-planning-control under latency and safety constraints; behavior trees/state machines vs learned policies; human-on-the-loop; geofencing, no-strike rules, mission abort; sim and field testing; ROS2/middleware patterns; sensor fusion; degraded modes; autonomy audit logging. Use for UAS/autonomous stacks, safety rules, HITL, sim-to-field validation, fail-safe—not LLM products (ai-engineer), LLM red team (ai-redteam), safeguard serving (ml-infrastructure-engineer-safeguards), governance only (ai-risk-governance), MCU firmware without autonomy (embedded-real-time-software-engineer), plant PLC/DCS (control-software-developer), HIL security bench (hardware-in-the-loop-security-tester).
Create, share, view, comment on, edit, and run human-in-the-loop review loops over markdown documents via Proof, the collaborative markdown editor at proofeditor.ai ("Proof editor"). Use when the user wants to render or view a local markdown file in Proof, share markdown to get a URL, iterate collaboratively on a Proof doc, comment on or suggest edits in Proof, HITL a spec/plan/draft for human review, sync a Proof doc back to local, or work from a proofeditor.ai URL. Trigger on phrases like "view this in proof", "share to proof", "iterate with proof", or "HITL this doc", and on ce-brainstorm / ce-ideate / ce-plan handoffs for human review. Also match clear requests for a rendered/shared markdown review surface even if the user does not name Proof. Do not trigger on "proof" meaning evidence, math/logic proof, burden of proof, proof-of-concept, or bare "proofread this" requests where inline text review is expected.
Drive external CLI coding agents (Claude Code, Codex, OpenCode) as first-class AG2 Agents over the Agent Client Protocol (ACP): point an Agent at a ClaudeCodeConfig / CodexConfig / OpenCodeConfig preset from ag2.acp and ask()/run() it like any agent, with its thinking, tool calls, plans and permission prompts externalized onto AG2's event stream. Use when orchestrating, observing, or gating CLI coding agents from Python — permission_policy ask/auto/deny HITL gating, fs_root confinement, turn timeouts, in-process testing via fake_acp_config. To expose an AG2 agent to other systems instead, see ag2-a2a or ag2-mcp; for approval plumbing see ag2-hitl.
UiPath Action Center human-in-the-loop tasks via `uip tasks`: list, assign, complete, plus task catalogs, comments, labels, metadata, and task data (get/save). For authoring HITL nodes in flows/agents→uipath-human-in-the-loop. For Orchestrator→uipath-platform, codedapp→uipath-coded-apps. Skip Document Understanding.
Work one session of an existing Decision map — load it, show the frontier (open, unblocked, unclaimed tickets) by name, claim exactly ONE, resolve it with the matching arc skill (grilling / prototype / research / task), record the answer on the ticket, graduate any fog it cleared through the dry-run gate, then STOP. This is also the skill that graduates fog on its own, when a map has nothing takeable left but fog remains. Use when a map already exists and the user says "continue the map", "next decision", "work the decision map", names a ticket on it, or comes back to a charted effort. Do NOT use to create a map or to add the first tickets to a foggy idea (that is chart-map), and do NOT use for a single-session design with no map behind it (that is grill-then-plan / sp-grill-with-doc). Never resolves more than one HITL ticket in a session.
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.
Design and implement autonomous AI marketing agent systems using the PRAL, BDI, and OODA frameworks. Invoke when a client is ready to move beyond reactive GenAI prompting to proactive, autonomous marketing workflows, or when planning an AI-first marketing operations architecture.
Use to summarize a recorded video via the LVS summarization microservice (HITL-gated) with a VLM fallback. Not for live RTSP captioning or incident-range reports.
Use to summarize a recorded video via the LVS summarization microservice (HITL-gated) with a VLM fallback. Not for report generation or live RTSP captioning.
Design WhatsApp LLM chatbots for East African markets: conversation flows, social presence principles, trust-building, local language registers, and human escalation protocols. Invoke when a client wants to automate WhatsApp customer service, sales enquiries, or support using AI.
Chart a Decision map for an effort too big for one agent session — name the destination, grill breadth-first to separate real decision tickets from fog, create the map and its tickets behind a dry-run gate, fire the research subagents, then STOP. Use when the user has a loose, foggy, multi-session idea — "this is huge, where do we even start", "plan this migration", "chart this", "make a decision map", "map out this initiative", "too big for one session" — and the route to the goal is not visible yet. Do NOT use for a well-scoped single-session design (that is grill-then-plan / sp-grill-with-doc), and do NOT use to continue a map that already exists (that is work-map). If the opening grill surfaces no fog, this skill stops and says a map is not needed.