Abe (TypeScript / Node ≥ 18) — npm install abe-ai

Abe is one control point before an AI agent acts.

It checks a proposed action against your policy and returns ACT, BLOCK or ESCALATE, plus an immutable, hash-verified Judgment-Grounded Record of why. It runs entirely on your machine: no account, no API key, no network, no model. Median evaluation time is well under a millisecond.

npm install abe-ai
npx abe init                      # writes abe-policy.yaml + request.json
npx abe check request.json

Decision: ESCALATE
Reason: FINANCIAL_THRESHOLD_EXCEEDED
Rules: purchase_review_limit
Risk: MEDIUM
Record: jgr_01M3Q4...

Python

from abe import Abe

abe = Abe("abe-policy.yaml")

result = abe.check(
    action={"type": "wire_transfer", "amount": 50000},
    context={"agent_id": "finance-agent"},
)

if result.decision == "ACT":
    ...                                   # execute exactly as evaluated
elif result.decision == "BLOCK":
    ...                                   # do not execute; tell the user result.reason_code
else:  # ESCALATE
    ...                                   # hold; ask a human, or let a resolver decide

result.record.to_dict()                   # the Judgment-Grounded Record (store it, ship it to your SIEM)

TypeScript

import { Abe } from "abe-ai";

const abe = new Abe({ policy: "./abe-policy.yaml" });
const result = await abe.check({
  action: { type: "purchase", amount: 12500, currency: "USD" },
  context: { agent_id: "procurement-agent", principal_id: "user_123" },
});
console.log(result.decision);             // "ESCALATE"

A policy

version: "0.1"
defaults: { unmatched: ESCALATE }           # nothing matched -> needs judgment (fail closed)
rules:
  - id: purchase_hard_limit
    when: { all: [ { field: action.type, op: eq, value: purchase }, { field: action.amount, op: gt, value: 100000 } ] }
    decision: BLOCK
    reason_code: HARD_POLICY_VIOLATION
  - id: purchase_review_limit
    when: { all: [ { field: action.type, op: eq, value: purchase }, { field: action.amount, op: gt, value: 10000 } ] }
    decision: ESCALATE
    reason_code: FINANCIAL_THRESHOLD_EXCEEDED
  - id: routine_purchase
    when: { all: [ { field: action.type, op: eq, value: purchase }, { field: action.amount, op: lte, value: 500 } ] }
    decision: ACT
judgment_required:
  - action.type: terminate_employee

Also: authorization per agent, required evidence, risk and irreversibility thresholds, confidence minimums. BLOCK beats ESCALATE beats ACT. Bad input or a broken rule returns ESCALATE / EVALUATION_FAILURE, never ACT. Full reference: SPEC.md.

Every way to run it

Library	pip install abe-ai · npm install abe-ai
CLI	abe check · validate-policy · validate-record · conformance · keygen
Local HTTP sidecar	abe serve --policy abe-policy.yaml → POST http://127.0.0.1:8787/v1/check (any language)
Docker sidecar	docker run -e ABE_TOKEN=… -v $PWD:/policy:ro -p 127.0.0.1:8787:8787 ghcr.io/flowinfosystems-index/abe
MCP server	pip install "abe-ai[mcp]" → abe mcp --policy /abs/abe-policy.yaml (tool: fjp_check_action)

Records you can audit

Every call returns a record with the decision, every matched rule, the exact policy hash, the request hash, and a SHA-256 record_hash (optionally Ed25519-signed with abe keygen). Records never change: outcomes and resolutions are appended as linked records. Each one is a valid FJP-CONF v0.1 Judgment-Grounded Record.

abe check request.json --record-out jgr.json --sign-key abe-signing-key.pem
abe validate-record jgr.json --public-key abe-signing-key.pub.pem

When rules aren't enough: Flow

Abe is deliberately useful without Flow. When it returns ESCALATE, you can hand that one action to Flow's judgment service and get a verb, a reason and a falsifiable record back:

from abe import Abe
from abe_flow import FlowResolver                     # pip install abe-flow

abe = Abe("abe-policy.yaml", resolver=FlowResolver(api_key=os.environ["FLOW_API_KEY"]))

ACT and BLOCK never leave your machine. Only escalations are sent (redacted), and if Flow is unreachable the answer simply stays ESCALATE. Human-approval escalations are never auto-resolved.

Conformance

abe conformance --bench        # FJP-CONF v0.1 Gate profile, Level 3 — 160 checks, offline

Repository

SPEC.md                  the normative specification
schemas/                 JSON Schemas (request, response, record, policy)
python/                  abe-ai (PyPI): core, CLI, HTTP, MCP, stores, signing, conformance
typescript/              abe-ai (npm): same API, same records, same hashes
flow-resolver/           abe-flow (PyPI): Gate → Flow for ESCALATE only
conformance/fixtures/    golden decisions + canonical-JSON vectors shared by every SDK
examples/                purchasing, software agent, physical agent, travel (+ Judd), Flow

Apache-2.0. FJP™, Abe™ (Abe) and FJP-CONF™ are trademarks of Flow Information Systems — see TRADEMARKS.md.

Notes for Node

• check() is async; resolvers may be sync or async and are bounded by resolverTimeoutMs (default 15 s).
• Records are deep-frozen objects: any mutation throws TypeError.
• Stores: MemoryStore, FileStore (JSON Lines), CallbackStore, or implement RecordStore (save/get/linked).
• Flow for Node: copy examples/flow/flow-resolver.mjs (zero dependencies).
• Records and hashes are byte-identical to the Python SDK (verified in CI).