Documentation

FORGE — the decision loop

Fetch → Orient → Resolve → Go → Extract

Every decision flows through this loop, creating a compounding record of organizational judgment.

Phase	What happens
Fetch	Load context and past decisions
Orient	Check guardrails and constraints
Resolve	Decide and record with reasoning
Go	Execute
Extract	Evaluate outcomes and distill patterns

Available as a Claude Code plugin with hooks, commands, and skills that automate the entire loop.

How it works

Two calls cover the full agent workflow:

Session start  → get_session_context  (cognitive context: profile, calibration, patterns)
Decision point → pre_action           (query + guardrails + record in one call)

The server auto-captures deliberation traces, bridge-definitions, and related decision links — zero client changes needed.

{
  "allowed": true,
  "similar_decisions": [],
  "guardrail_results": [],
  "calibration_context": { "brier_score": 0.024, "accuracy": 0.963 },
  "decision_id": "abc123"
}

Start here

Getting StartedInstall with Docker, run locally, or wire it in as a skill.

Golden Path WalkthroughEnd-to-end: query, check, record, review, calibrate.

MCP IntegrationConnect Claude Code, Claude Desktop, and other MCP clients.

API ReferenceCSTP over JSON-RPC 2.0 — every method, parameter, and response.

Core capabilities

Query before deciding

Semantic search across past decisions. Find what worked, what failed, and why — by structure (“where did we use this pattern?”) or function (“what solved this problem?”).

Guardrails

Policy enforcement that prevents violations before they occur. Block high-stakes decisions with low confidence. Require review for production changes.

Deliberation traces

Every query and guardrail check is automatically linked to the resulting decision — full provenance of how each choice was made.

Calibration & analytics

Track Brier scores, success rates, and confidence calibration over time. Reason-type analytics show which reasoning patterns predict success.