nanoclaw

Generate or edit images with Google Gemini image models (Nano Banana 2 / Pro) and send them to the chat. Use when the user asks for an image, a creative, an ad variant, a mockup, an infographic, or an edit of an uploaded picture.

Meta Ads reporting, monitoring, and optimization. Read campaign performance, detect anomalies, identify waste, recommend actions. Use when the user asks about ad performance, wants reports, or mentions Meta/Facebook ads.

Add pond cross-session recall. Agents search their own past sessions (and any stores the operator grants) via read-only MCP tools, with write (ingest) and read permissions controlled independently per store. Use when the user wants agents to remember past conversations, search session history, or recover context after compaction.

Search your own past sessions (and any additionally granted pond stores) with the pond MCP tools. Use when you need context from an earlier conversation, a past decision, or anything that happened beyond your current context window.

Add clidash — a zero-dependency, read-only web dashboard that derives its tabs and tables at runtime from any CLI that lists resources as JSON. Ships pre-wired for NanoClaw's ncl CLI (agent groups, sessions, channels, users, roles), plus message-activity charts, a log tail, and a read-only file viewer for group skills/CLAUDE.md/profiles.

Speak: turn text into a voice note and send it as a Telegram voice bubble. Use this when the user asks you to "say", "speak", "read aloud", "send a voice message / voice note / voice reply", or when a spoken reply is clearly better than text (a short greeting, a poem, something emotional, an accessibility need). Uses Google Gemini TTS. Stay text by default — only speak when asked or when voice genuinely improves the moment.