truceptor 0.5.0

Truceptor — trusted intercept for agent tool calls: allow/deny/escalate (does not execute tools).

Truceptor (Python SDK)

Truceptor — trusted intercept for agent tool calls: preflight allow/deny/escalate; your runtime executes tools after allow.

CLI: acp up, acp init (also truceptor up). Install: pip install truceptor.

Import: from truceptor import governed_tool, ACPClient. Env vars use ACP_* (ACP_INTERCEPTOR_URL, etc.). User data dir stays ~/.acp/.

Add governance, approvals, policy enforcement, and execution visibility to AI agents in minutes.

Works with CrewAI, LangGraph, Strands, Google ADK, MCP tools, and custom Python workflows.

Requires: Docker Desktop (Compose v2).

Full integration guide: docs/INTEGRATION.md (setup vs runtime, Rego → Postgres, pip paths, who owns what).

---

Quick start

Make sure Docker Desktop is already running before you start Truceptor (acp up).

pip install truceptor
acp init          # ~/.acp/policies + ~/.acp/agents.json
acp up            # Open mode (default)
acp dashboard

Mode	Command
Open (default)	acp up
Seeded	acp up --register
Enforced	AGENT_REGISTRY_ENFORCE=true acp up
Check mode	acp mode

Step	Command
Install	pip install truceptor
Initialize policies	acp init
Start stack	acp up
Open UI	acp dashboard → http://localhost:3090/dashboard/
Stop	acp down

For a real local integration, the usual sequence is:

1. acp init
2. start Docker Desktop
3. acp up
4. export ACP_INTERCEPTOR_URL=http://localhost:8080
5. export ACP_BEARER_TOKEN=...
6. run your governed agent

If Docker is not already running, acp up will fail with Docker daemon / compose connection errors.

What acp init and acp up actually do

Step	What you get
acp init	Creates ~/.acp/policies/ with starter Rego (router.rego, rbac.rego, …) and HOW_TO_ADD_RULES.md. You edit these files — Truceptor does not generate rules from your agent code.
acp up	Starts Postgres, OPA, interceptor, dashboard. On startup the interceptor loads your ~/.acp/policies, upserts them into Postgres, and pushes them to OPA.
Your agent	Sends only agent_id, tool, action, and context to POST /tool-call — never Rego text.

After changing any .rego file: acp down && acp up. The dashboard Policies tab is read-only and shows plain-English rule summaries (Rego bundle and Policy Studio policies). Business users can author rules in the Policy Studio tab (Save & activate); compiled Rego lands in the same Postgres policy_versions pipeline.

Policy Studio (dashboard)

Step	What you get
Policy Studio → New rule	Wizard: rule type → configure thresholds → simulate → Save & activate
Activate	Compiler writes studio/*.rego + studio_router.rego into policy_versions and syncs OPA
Policies tab	Same catalog API — studio policies tagged Policy Studio, engineer modules tagged Rego bundle

See docs/POLICY_STUDIO.md.

---

Where your policies live (pip install)

The Python package installs into your virtualenv (.venv). Your editable Rego files do not — they live in your home directory:

What	Location
Installed package (read-only)	.venv/.../site-packages/truceptor/
Your policy folder	~/.acp/policies/ (created by acp init)
Dev JWT keys	~/.acp/jwt/
Docker runtime bundle	~/.acp/runtime/<version>/

acp init copies starter templates from the package into ~/.acp/policies/. You edit those files for your tools, roles, and allow/deny/escalate rules.

acp up bind-mounts ~/.acp/policies into the interceptor container. On startup Truceptor upserts them into Postgres, then pushes to OPA. Nothing is uploaded from your agent at runtime.

Use a client repo instead of ~/.acp/policies:

export ACP_POLICIES_DIR=/path/to/my-project/policies
acp init && acp up

Two phases: (1) Setup — write Rego, run acp up; (2) Runtime — SDK sends agent_id, tool, action only. See the integration guide.

---

Why Truceptor?

Most agents call tools, APIs, and other agents directly. Teams then scatter rules across Python, workflows, and frameworks:

if supplier_risk_score > threshold:
    require_human_approval()

That becomes inconsistent, hard to audit, easy to bypass, and duplicated everywhere.

Truceptor centralizes governance outside agent code:

Capability	What you get
Centralized governance	One place for rules, not copy-paste per team
Policy enforcement	OPA/Rego evaluates every governed call
Approvals	Escalate high-risk actions to humans
A2A governance	Governed agent-to-agent calls
A2T governance	Governed agent-to-tool calls
Audit & visibility	Decisions, traces, registry in one dashboard

---

Architecture

Agent runtime -> Truceptor SDK -> Interceptor / Gateway -> OPA / Rego -> allow | deny | escalate -> governed execution

Policy lifecycle (startup): PostgreSQL policy_versions -> interceptor -> PUT /v1/policies -> OPA in-memory cache
Runtime audit: decisions and traces -> PostgreSQL; dashboard reads catalog and forensics from interceptor APIs

SETUP (acp init + acp up)
  ~/.acp/policies/*.rego  ──mount──►  interceptor container
        │
        ▼
  Upsert → Postgres (policy_versions) → push → OPA

RUNTIME (every tool call)
  Agent + SDK  ──POST /tool-call──►  Interceptor  ──►  OPA  ──►  allow | deny | escalate
        (agent_id, tool, action — no Rego in request)

Truceptor sits at the execution boundary:

• agent frameworks keep reasoning and orchestration
• PostgreSQL is the source of truth for Rego policies, audit events, agents, and approvals
• on startup the interceptor loads your ~/.acp/policies plus built-in platform samples, upserts to Postgres, then syncs to OPA
• the interceptor validates identity, routes policy, and records audit traces
• OPA / Rego returns deterministic allow, deny, or escalate decisions against synced rules
• only allowed or approved requests reach tools, APIs, MCP servers, or other agents

Request sequence

Typical governed flow:

1. On interceptor startup: load active Rego from Postgres and sync to OPA via the Policy API
2. the agent calls a governed function through the Truceptor SDK
3. the interceptor validates JWT identity and normalizes the request
4. OPA / Rego evaluates policy for the tool, action, and claims (in-memory rules synced from Postgres)
5. Truceptor records the decision in Postgres and exposes it in the dashboard
6. the SDK executes only on allow, waits for review on escalate, and blocks on deny

---

Dashboard

The Truceptor governance dashboard is a core differentiator: live allow/deny/escalate decisions, approvals, agent registry, and policy catalog. Open http://localhost:3090/dashboard/ after acp up.

Overview & activity

Decisions & approvals

Registry & policies

Forensics & replay

---

Deployment modes

Mode	How	Best for
Local	acp up via pip + Docker	Demos, dev, quickstart
SDK	@governed_tool in your agent code	CrewAI, LangGraph, Strands, custom Python
Gateway	Single origin on :3090 (dashboard + API proxy)	Local unified URL; pattern for prod ingress
Cloud / self-hosted	Docker Compose, Kubernetes, ECS/EKS on AWS/Azure/GCP	Team or enterprise rollout

Local endpoints

URL	Purpose
http://localhost:3090/dashboard/	Governance dashboard
http://localhost:8080	Interceptor API (/tool-call, /api/v1/*)

Self-hosted (example)

https://acp.your-company.example

---

Example: governed tool

Set ACP_INTERCEPTOR_URL and ACP_BEARER_TOKEN before defining a @governed_tool.

The decorator constructs its client at decoration time, so setting those env vars later inside main() is too late.

import os
from truceptor import governed_tool

# Set env vars before decorator evaluation.
os.environ.setdefault("ACP_INTERCEPTOR_URL", "http://localhost:8080")
os.environ.setdefault("ACP_BEARER_TOKEN", "<dev-jwt>")

@governed_tool(
    agent_id="texas-weather-agent",
    tool="weather_api",
    action="read_weather",
)
def get_texas_weather(city: str):
    return {"city": city, "status": "pending_review"}

Truceptor intercepts the call, evaluates policy, then allows, denies, or escalates.

tool="weather_api" is the governed tool id. It does not have to match the Python function name (get_texas_weather). Policy keys on tool, not the Python symbol name.

The JWT agent_id claim must match agent_id="texas-weather-agent" in @governed_tool(...).

---

Example: policy (Rego)

package acp.policy

allow {
    input.identity.role == "supply-chain-manager"
    input.action.tool == "supplier_approval"
    input.risk_score < 70
}

Edit policies in ~/.acp/policies/ after acp init, then restart with acp down && acp up so the interceptor upserts them into Postgres and syncs to OPA. See HOW_TO_ADD_RULES.md in that folder after init.

---

Adding a new governed tool

acp init creates starter policy files in ~/.acp/policies/, but you still need to wire new tools into policy yourself.

For a brand-new governed tool, update all of these:

1. choose a governed tool id, for example weather_api
2. add tool-to-role mapping in ~/.acp/policies/rbac.rego
3. route the tool in ~/.acp/policies/router.rego
4. add or reuse a domain policy file such as weather.rego
5. mint and export ACP_BEARER_TOKEN
6. make sure the JWT agent_id claim matches @governed_tool(agent_id=...)
7. restart: acp down && acp up (loads Rego into Postgres + OPA)
8. run the agent

Two valid patterns:

Pattern 1: reuse a governed tool bucket

Use one tool id for multiple Python functions in the same policy domain.

@governed_tool(
    agent_id="texas-weather-agent",
    tool="weather_api",
    action="read_weather",
)
def get_texas_weather(city: str):
    ...

~/.acp/policies/rbac.rego

"weather_api": ["weather"],

~/.acp/policies/router.rego

decision = weather.decision {
    input.tool == "weather_api"
    rbac.decision.decision == "allow"
}

~/.acp/policies/weather.rego

package acp.weather

default decision = {
    "decision": "allow",
    "reason": "weather_action_allowed",
    "policy": "acp/weather",
}

decision = {
    "decision": "deny",
    "reason": "weather_action_not_permitted",
    "policy": "acp/weather",
} {
    not input.action == "read_weather"
    not input.action == "read_weather_batch"
}

Pattern 2: add a brand-new governed tool id

If you want policy to key directly on a new id such as get_texas_weather, add that id to both rbac.rego and router.rego, then route it to the right domain policy file.

If your starter router.rego or rbac.rego references a tool id, make sure the corresponding domain policy file also exists.

---

Local JWT dev example

Local governed /tool-call requests require a bearer token when JWT auth is enabled. ACP_INTERCEPTOR_URL alone is not enough.

Set:

• ACP_INTERCEPTOR_URL=http://localhost:8080
• ACP_BEARER_TOKEN=<signed-jwt>

If ACP_BEARER_TOKEN is missing, governed calls may fail with 401 Unauthorized / missing bearer token.

One dev-friendly way to mint a token is:

python -m pip install pyjwt cryptography

export ACP_BEARER_TOKEN="$(
python - <<'PY'
from pathlib import Path
import time
import jwt

private_key = (Path.home() / ".acp" / "jwt" / "private.pem").read_text()
token = jwt.encode(
    {
        "sub": "agent:texas-weather-agent",
        "agent_id": "texas-weather-agent",
        "roles": ["weather"],
        "iss": "acp-dev",
        "aud": "acp-interceptor",
        "exp": int(time.time()) + 3600,
    },
    private_key,
    algorithm="RS256",
)
print(token)
PY
)"

Use these claims for local development:

• sub=agent:texas-weather-agent
• agent_id=texas-weather-agent
• roles=["weather"]
• iss=acp-dev
• aud=acp-interceptor

agent_id in the JWT must match the agent_id passed to @governed_tool(...).

---

Example: governance flow

Supply Chain Agent
    → Truceptor validates identity (JWT)
    → OPA evaluates policy
    → Decision: ESCALATE
    → Human approves in dashboard
    → Execution resumes

---

What Truceptor provides

• Policy enforcement — OPA/Rego (Cedar on roadmap)
• Identity — JWT from Okta, Auth0, Keycloak, or your IdP
• Approvals — human-in-the-loop for risky actions
• Observability — dashboard for decisions, traces, agents, tools
• Agent registry — lightweight catalog of agents and capabilities
• Framework-friendly — keep CrewAI / LangGraph / Strands for reasoning; govern execution here

---

Philosophy

Orchestration frameworks handle reasoning, planning, and workflows.

Truceptor handles governance, trust, approvals, policy, and auditability.

Reasoning stays autonomous. Execution stays governed.

---

Roadmap

• Policy Studio — MVP shipped in 0.5.0: Approval, Risk, and Compliance threshold rules via dashboard wizard; compiles to Rego in policy_versions; human-readable Policies tab for all modules.
• Access rules and agent-to-agent rules (Policy Studio)
• Enterprise topology views
• Multi-cloud deployment templates

---

Release notes

0.5.0

• Breaking (Python import): package directory renamed acp → truceptor. Use from truceptor import governed_tool, ACPClient. CLI unchanged (acp up, acp init); env vars still ACP_*; user dir still ~/.acp/.
• Policy Studio MVP in bundled interceptor + dashboard: business rules → Rego → same Postgres/OPA pipeline.
• Policies catalog: human-readable rule summaries for engineer-authored and Policy Studio modules.
• Bundled runtime sync excludes monorepo-only deployments/ dev scripts from the PyPI wheel.

0.4.x

• PyPI package truceptor; integration docs; bundled Docker runtime via acp up.

---

License

MIT License