Capabilities · Built on Claude

What GRID runs on Claude today

One row per capability, one production surface where it lives, one honest sentence about how. Everything marked Live is running in the product now; Staged means the scaffolding is wired and the optimization is in progress.

Long-context reasoning (≥ 100k tokens per call)

Weekly narrative · /environments/[slug]
Live

The 5-sentence Monday memo composes 7 days of AuditLog + signals + goal deltas into one Claude call. One round trip, one memo. Cached 24h.

Sub-agent dispatch (specialist per task)

Project executor chain · /projects/[id]
Live

A Project plan fans out to 3–10 specialist executors — Notion fetch, Claude draft, Canva design, Gmail draft. Each gets scoped context. Auto-runs until a human gate.

Tool use with structured schemas

Skill registry · /skill-space
Live

14 real tool adapters and a four-dimensional skill classifier (Location, Action, Interaction, Execution). Claude selects from a menu and emits a JSON plan.

Click any Atrium action in the ledger. Drawer shows what Atrium read, what tool it called, what it produced, tokens and cost. KernelTrace + IntelligenceLog, no black boxes.

Persistent memory per organization

Atrium Memory + Mastery Insights · /memory
Live

AtriumMemory holds user-taught corrections; MasteryInsight holds patterns Atrium derived from execution reviews; OperationalPlaybook holds the regenerated "how work actually runs" doc. All org-scoped.

Bidirectional teaching loop (user ↔ model)

Atrium Academy · /learn
Live

Four fluency capabilities (delegation, review, context-giving, trust-calibration) scored from aggregates. Daily lesson biased to the weakest. Every answer writes a AtriumMemory entry future Atrium calls RAG against.

Streaming build narration (SSE)

Onboarding build step · /welcome
Live

Atrium streams scaffold progress as it creates the System and Workflow. Real steps, real writes, rendered one line at a time.

Reversible autonomous actions

24-hour undo window · /environments/[slug]
Live

Every autonomous action gets an AuditLog undo path and a AtriumMemory user_correction on reversal. Future Atrium calls factor in the correction.

Calibrated autonomy (5 levels)

AutonomyConfig per Workflow · /approvals
Live

Observe → Suggest → Act & Notify → Autonomous → Self-Direct, per Workflow and per System. Trust Score recalibrates from approval rate.

Override reason capture (structured)

Teach Atrium pill row · /environments/[slug]
Live

Three override categories plus one-line context, per rejected action. Written to AtriumMemory as type=user_correction with category set.

Prompt caching (cost shape)

Narrative + planner calls · server-side
Staged

System prompts are structured so the static prefix is cache-friendly. Cache adoption is a continuing optimization, flagged here so partners know we track it.

Staged

Every tool adapter implements a single Executor interface (step, project) → ExecutorResult with artifacts, trace, mode. Maps cleanly to MCP when we expose it externally.

The architecture brief that expands every row lives at /security/architecture. The roadmap for what comes next lives at /roadmap.