session #00417 · live
A memory stronghold for your AI coding agent.
Deploy once. Every session your team's agents have — Claude Code, Codex, Copilot, Kiro — flows into one Postgres knowledge graph. Next session, every agent walks in already briefed.
Enterprise deployment. We handle the rollout — Postgres provisioning, Entra JWT wiring, namespace setup, and CLAUDE.md bootstrap. Contact us to get started.
Contact us ↗What Borg knows about this project, right now
LIVE GRAPH NEIGHBORHOOD · ns=product-engineering
entity fact (typed predicate) episode
TASK SUCCESS — ENGINEERING-TASK BENCH10/ 10Borg-compiled context completes every task. No-memory baseline: 0/10.
RETRIEVAL PRECISION91.3%+12.7 pts over top-10 vector RAG. Measured against labeled ground-truth.
STALE-FACT RATE−78%Temporal supersession keeps contradicted decisions out of compiled context.
◆borg_think · classify → retrieve → rank → compile◆postgres-native · pgvector · recursive CTE◆multi-profile retrieval · reciprocal rank fusion◆24 canonical predicates◆3-pass entity resolution◆7 MCP tools · think · learn · recall · fetch · delete · guard · summarize◆claude code · codex · copilot · kiro◆valid_from / valid_until supersession◆entra JWT · RBAC · namespace isolation◆15 tables · 1 function◆78% fewer stale facts vs RAG◆borg_think · classify → retrieve → rank → compile◆postgres-native · pgvector · recursive CTE◆multi-profile retrieval · reciprocal rank fusion◆24 canonical predicates◆3-pass entity resolution◆7 MCP tools · think · learn · recall · fetch · delete · guard · summarize◆claude code · codex · copilot · kiro◆valid_from / valid_until supersession◆entra JWT · RBAC · namespace isolation◆15 tables · 1 function◆78% fewer stale facts vs RAG
§ 01 — THE LOOP
Not search.
Compilation.
Every borg_think call runs the same five-stage pipeline: classify the intent, retrieve across facts / episodes / graph, rank on four dimensions, compile for the target model, log the trace. You see every candidate and why it was picked or rejected.
Illustrative trace · claude-code · ns: product-engineering
task: debuggraph_neighborhoodepisode_recallrrf_merge
Classify intent · debug
Retrieve across 2 profiles
Fuse with RRF, rank 14, select 6
Compile XML for claude
Audit log + access tracking
§ 02 — WHAT'S INSIDE
Six decisions,
one system.
Every capability below ships in the commercial Borg enterprise deployment. Read the full tradeoffs on /features.
01
Knowledge Graph Extraction
An LLM pipeline turns conversations into typed entities, temporal facts, and reusable procedures — with three-pass resolution that prefers fragmentation over collision.
↳ Graph is mergeable. Collisions are not.
MODEL · gpt-5-mini
CAP · 10 entities / ep.
THRESH · 0.92 semantic
02
Temporal Facts with Supersession
Every fact carries valid_from and valid_until. New contradictions mark old facts superseded — never deleted. Compliance queries still see history; working memory doesn't.
↳ 78% fewer stale facts in bench C vs RAG.
STATUSES · 7
PREDICATES · 24 canonical
HISTORY · bitemporal
03
Task-Specific Compilation
Dual-profile intent classification picks what kind of memory a query needs. Debug gets episodic + procedural. Architecture gets semantic facts. Compliance excludes procedural entirely.
↳ Weights bias ranking, not hard excludes.
PROFILES · 2
FORMATS · XML · JSON
BUDGET · per namespace
04
Inspectable Ranking
Four orthogonal scores — relevance, recency, stability, provenance — surfaced individually. Every compilation emits an audit trail of selected, rejected, and why. No opaque composites.
↳ The audit log is the debugger.
DIMENSIONS · 4
AUDIT · borg_audit_log · pgAudit
OPS · compile trace · write · DDL
05
Namespace Isolation
Hard by default: every entity, fact, and episode belongs to exactly one namespace. Core memory retrieval is namespace-scoped. Per-namespace token budgets keep projects separate.
↳ Tenant-wide inspection requires explicit admin permission.
MODE · hard
BUDGETS · token-per-ns
ACCESS · group · permission checked
06
PostgreSQL Maximalism
One database. Graph via recursive CTE. Embeddings via pgvector. Audit via pgAudit. No Qdrant. No Neo4j. No sync daemons. Runs on Supabase, Neon, managed Postgres, or your laptop.
↳ Nothing gets out of sync if there is nothing to sync.
TABLES · 15 + 1 fn
EXT · pgvector · pgAudit
MIN · PG 14
§ 03 — MCP SURFACE
Seven tools.
Full surface.
Seven MCP tools covering the full write/read/delete/guard surface. Works unchanged across Claude Code, Codex CLI, Copilot, Kiro, and anything else that speaks MCP over streamable HTTP.
borg_thinkcompile
Compile context for a query. Runs the full online pipeline — classify, retrieve, rank, compile — and returns a ranked, token-budgeted context package.
borg_think(
query: "debug webhook delivery timeout",
namespace: "product-engineering",
model: "claude",
task_hint: "debug"
)
borg_learnrecord
Record a decision or discovery. Stored immediately, extraction runs async. Returns in milliseconds so nothing blocks the agent loop.
borg_learn(
content: "Versioned event payloads via schema registry…",
source: "claude-code",
namespace: "product-engineering"
)
borg_recallbrowse
Direct memory search without compilation. Raw episodes, facts, and procedures — for UI, audits, or you-the-developer.
borg_recall(
query: "release checklist",
namespace: "product-engineering",
memory_type: "semantic"
)
borg_get_episodefetch
Fetch a specific episode by ID. Returns the full stored content and metadata. Useful for audit trails and deep-linking into prior sessions.
borg_get_episode(
episode_id: "11111111-1111-1111-1111-111111111111"
)
borg_forgetdelete
Soft-delete episodes, facts, entities, or procedures. Marks records deleted with reason and timestamp; episode deletion can cascade to derived memory.
borg_forget(
target_type: "episode",
target_id: "11111111-1111-1111-1111-111111111111",
namespace: "product-engineering",
reason: "contains sensitive credentials"
)
borg_guardguard
Risk-assess a proposed action against stored facts. Returns risk level, contradicting facts, and recommended decision. Prevents the agent from repeating known-bad moves.
borg_guard(
action: "drop table events",
namespace: "product-engineering",
proposed_change: "remove the events table in migration 42"
)
borg_project_summarysummarize
Generate a deterministic project summary from stored facts, decisions, and episodes. Stable output for CLAUDE.md bootstrapping, onboarding, and context priming.
borg_project_summary(
namespace: "product-engineering"
)
§ 04 — THE LANDSCAPE
Everyone else returns results.
Borg returns context.
Every competitor wraps 2–3 databases and hands the agent raw matches. Borg is one Postgres that compiles.
| Mem0 | Zep | Cognee | Borg | |
|---|---|---|---|---|
| Databases required | Qdrant + Neo4j | Neo4j + vectors | 3+ | 1 · Postgres |
| Integration | Python SDK | Python SDK | Python SDK | 1 line in CLAUDE.md |
| Retrieval output | Raw results | Raw results | Raw results | Compiled ranked context |
| Stale-fact handling | None | Temporal (Neo4j) | None | Temporal · supersession |
| Cross-client reuse | Per SDK | Per SDK | Per SDK | Any MCP client, same DB |
| Audit + inspect | Limited | Partial | Limited | Full trace, per compile |
| License | Mixed | Mixed | Open | Commercial |
§ 05 — THE BENCH
Ten tasks.
Three conditions.
Modeled on real cloud-infrastructure engineering work — incident response, schema migration, architecture review. Each condition runs the same evaluator against the same ground-truth labels. Seeds + code: /bench.
Condition A
No memory
Baseline — every session starts from zero.
Task success0/ 10
Retrieval precision6.0%
Stale-fact rate0.0%
Condition B
Top-10 vector RAG
Standard similarity search over conversation chunks.
Task success8/ 10
Retrieval precision81.0%
Stale-fact rate11.5%
Condition C
Borg · compiled
Full pipeline: classify, retrieve, rank on 4 dims, compile.
Task success10/ 10
Retrieval precision91.3%
Stale-fact rate2.5%
§ 06 — GO DEEPER
Three long reads.
One system.
Written for staff engineers and eng leaders. No buzzwords, no emoji, no filler.
/features
Features, with tradeoffs.
Every capability plus the decision it implies. Read why fragmentation beats collision, how supersession settles contradictions, and where the 24 predicates come from.
open ↗/architecture15 tables. One schema.
Reference layout, end-to-end schema, deployment topology, failure modes. Every table separates canonical data from serving state.
open ↗/scienceWhy this works.
Grounded in Tulving's episodic/semantic split, Fellegi-Sunter record linkage, bitemporal data, and the Lost-in-the-Middle paper. Memory is not a search problem.
open ↗— install it now —
One command. One Postgres.
Every session after, smarter.
Enterprise-ready. Entra JWT auth, RBAC, audit logging, namespace isolation. Runs on any managed Postgres.
Enterprise deployment. We handle the rollout — Postgres provisioning, Entra JWT wiring, namespace setup, and CLAUDE.md bootstrap. Contact us to get started.
Contact us ↗