CLI commands

pond <command> --help   # full flags for any command

Setup

• pond init - guided setup-and-repair: storage, adapters, MCP + agent skill, schedule
• pond adapters list|discover|enable|disable - manage which adapters sync ingests
• pond storage check|use - probe a destination; switch the active store
• pond creds add|list|delete - manage URL-scoped credential sets
• pond schedule start|stop|status|logs - manage the automatic sync schedule
• pond config show|path|schema - inspect resolved configuration

Data

• pond sync [<adapter>] - import, embed inline, fold indexes; --dry-run preview, --no-wait skip if locked, --verify reconcile, --format json
• pond optimize - embed the backlog, fold the indexes; --only / --skip <stage>
• pond copy --from <x> --to <y> - move data between stores, .pond archives, JSONL; --verify-only

Query

• pond search <query> - search (--mode vector|fts, --sort-by relevance|recency); --explain for the query plan
• pond get-session <ID> - fetch a session
• pond get-message <ID> - fetch a message
• pond sql <query> - one read-only SQL query over the corpus
• pond status - health, row counts, embedding coverage, schedule state; --hosts fleet view, --include-subagents count subagents separately

pond status --format json

One machine-readable document for scripting. Top-level keys: initialized (whether the store has been set up), pond_version (the producing binary's version, so consumers can pin the format), storage, corpus, indexes, embedding, source_agents (count of distinct source agents in the store), local (this host's adapters + pending-sync counts, last sync, next scheduled run), hosts, and schedule ({active, backend, every}).

Serve

• pond serve --transport http|stdio - run the API (HTTP default)
• pond mcp - alias for pond serve --transport stdio
• pond completions <shell> - shell completion scripts
• pond skill - print the bundled agent-onboarding SKILL.md

Copy endpoints

local local default dir · @ configured store · *.pond archive · *.jsonl or - JSONL wire stream.