VelesDB

Compress an agent's working context deterministically with VelesDB's context compiler before sending it to a model — fewer tokens, same facts, every decision auditable and reversible. Use when a prompt is bloated with repeated context, long logs, or accumulated conversation turns; when token costs need to drop measurably; or when the user says "compress my context", "optimize my prompt", "reduce token usage", "context is too big", or asks why part of their context was dropped; or when a session should be resumable later (save the working context at the end, load it back at the start of the next one). Requires the velesdb-memory MCP server (tools: compile_context, compile_transcript, context_savings, explain_compilation, retrieve_context_source, save_working_context, load_working_context).

Use durable, explainable, self-improving memory across a coding session via the velesdb-memory MCP server. Trigger whenever the velesdb-memory MCP tools (remember/recall/recall_fused/relate/why/feedback/forget) are available and the work would benefit from remembering decisions, recalling prior context, or answering "why did we do X". Use it at the START of a task (recall what's known), when a decision or durable fact emerges (remember + relate it), when asked why something is the way it is (why), and after using a recalled memory (feedback). Especially relevant for: multi-session projects, architecture/config decisions, incident postmortems, "why is this value/setting like this", onboarding to an unfamiliar codebase, and any place a fact learned now must survive to a later session. Do NOT use for transient chatter, secrets, or one-off scratch notes.

Re-run the VelesDB core-vs-ecosystem gap analysis (code + documentation). Use when asked to audit API parity, check that every crate/SDK/integration/doc has caught up to velesdb-core, verify the core↔children architecture is clean, or produce a parity matrix. velesdb-core is the single source of truth; the VelesDB Core License boundary is enforced in every recommendation.

Compress an agent's working context deterministically with VelesDB's context compiler before sending it to a model — fewer tokens, same facts, every decision auditable and reversible. Use when a prompt is bloated with repeated context, long logs, or accumulated conversation turns; when token costs need to drop measurably; or when the user says "compress my context", "optimize my prompt", "reduce token usage", "context is too big", or asks why part of their context was dropped. Requires the velesdb-memory MCP server (tools: compile_context, context_savings, explain_compilation, retrieve_context_source).