local-first · open source · production ready
SEAM is a local-first memory runtime for AI agents. Compile source into canonical MIRL records, persist in SQLite, derive retrieval indexes, and emit token-bounded context — through CLI, MCP, REST, and dashboard.
Operators and AI agents interact through CLI, MCP, REST, or the browser dashboard. All surfaces call the same shared runtime — no reimplementation, same guarantees everywhere.
Every piece of data follows a strict pipeline from untrusted source to agent-ready context. SQLite is canonical truth. Vector stores are rebuildable acceleration. PACK is derived, never authoritative.
SEAM separates concerns into explicit layers, each with its own engineering guarantees and contracts.
Preserve source phrasing and exact evidence. Source identity, hashes, exact spans, provenance linkage, and prompt-injection containment.
Canonical machine-readable semantic representation. Deterministic record shape, stable identities, entity consistency, uncertainty, and temporal semantics.
Dense, token-bounded retrieval projection. Relevance ranking, token budgets, reference retention, and explicit lossy vs exact behavior.
Shape canonical records for a task, operator, or interface. View-specific filtering with stable references back to canonical records.
SEAM ranks evidence through lexical, vector, graph, temporal, hybrid, and mix retrieval modes. Measure displacement and precision — not recall alone. The mix mode combines all signals for maximum context quality.
Every component is testable, every claim is auditable, and every index is rebuildable.
MIRL records persist durably in SQLite. Vector stores are acceleration layers — never truth. Everything derived can be rebuilt.
Rank evidence through lexical, vector, graph, temporal, hybrid, and mix signals. Traces included.
PACK creates dense, budget-aware projections. Reference retention, provenance fallback, explicit lossy vs exact.
Standard MCP stdio protocol. Works with Gemini, Claude, Cursor, and any MCP-compatible client.
FastAPI/Uvicorn server with auth, CORS, rate limiting, SSRF controls, and a full browser dashboard.
Source-to-MIRL transformation with provenance preservation. Evidence survives compilation.
Entity relationships, temporal semantics, contradiction state. Supersession-aware ranking prevents stale data.
Retrieved content never gains authority. Prompt-injection containment, scope isolation, atomic writes.
Every interface calls shared runtime behavior. No reimplementation. Same guarantees.
Operator-facing composition surface.
seam <command>Standard agent-tool protocol via stdio.
stdio · agentsFastAPI with browser dashboard.
seam serve :8765Operator observation and control.
seam dashboardThese are not aspirational — they are enforced.
If you find SEAM valuable, consider supporting the research.