Metadata-Version: 2.1
Name: agentic-sidecar
Version: 0.6.1
Summary: Companion intelligence and real-time decision supervision for autonomous AI agents.
Project-URL: Homepage, https://github.com/DeepAgentLabs/agentic-sidecar
Project-URL: Repository, https://github.com/DeepAgentLabs/agentic-sidecar
Project-URL: Issues, https://github.com/DeepAgentLabs/agentic-sidecar/issues
Project-URL: Changelog, https://github.com/DeepAgentLabs/agentic-sidecar/blob/main/CHANGELOG.md
Author: agentic-sidecar Contributors
License: MIT License
        
        Copyright (c) 2026 pramodbn27
        
        Permission is hereby granted, free of charge, to any person obtaining a copy
        of this software and associated documentation files (the "Software"), to deal
        in the Software without restriction, including without limitation the rights
        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
        copies of the Software, and to permit persons to whom the Software is
        furnished to do so, subject to the following conditions:
        
        The above copyright notice and this permission notice shall be included in all
        copies or substantial portions of the Software.
        
        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
        SOFTWARE.
License-File: LICENSE
Keywords: agents,decision-supervision,governance,intent,llm,safety
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Python: >=3.10
Requires-Dist: pydantic<3,>=2.0
Requires-Dist: pyyaml<7,>=6.0
Requires-Dist: rich<14,>=13.0
Requires-Dist: typer<1,>=0.9
Provides-Extra: agentic-chaos
Requires-Dist: agentic-chaos>=0.1; extra == 'agentic-chaos'
Provides-Extra: agenticlens
Requires-Dist: agenticlens>=0.1.3; extra == 'agenticlens'
Provides-Extra: anthropic
Requires-Dist: anthropic<2,>=0.49; extra == 'anthropic'
Provides-Extra: dev
Requires-Dist: build>=1.2; extra == 'dev'
Requires-Dist: mypy>=1.10; extra == 'dev'
Requires-Dist: pytest-cov>=5.0; extra == 'dev'
Requires-Dist: pytest>=8.0; extra == 'dev'
Requires-Dist: ruff>=0.6; extra == 'dev'
Requires-Dist: twine>=5.1; extra == 'dev'
Requires-Dist: types-pyyaml>=6.0; extra == 'dev'
Provides-Extra: docs
Requires-Dist: mkdocs-material>=9.0; extra == 'docs'
Provides-Extra: langgraph
Requires-Dist: langchain-core<2,>=0.3; extra == 'langgraph'
Requires-Dist: langgraph<2,>=1.0; extra == 'langgraph'
Provides-Extra: openai
Requires-Dist: openai<3,>=1.68; extra == 'openai'
Description-Content-Type: text/markdown

# agentic-sidecar

**A companion intelligence and real-time supervision layer for autonomous AI agents.**

> The Main Agent acts. The Sidecar observes, thinks, advises, and governs.

## Status

**Current checkout version: 0.6.1 (release preparation).** The runtime brings together deterministic
intent/policy/risk supervision, optional evaluators, budget accounting,
narration, and the CLI. See the [roadmap](ROADMAP.md) for release history
and upcoming capability stages.

| Capability | Current behavior |
| --- | --- |
| Runtime, policy/risk, Intent Guardian | Deterministic rules, constraint bindings, expiry warnings, intent/history injection |
| Observe / Govern modes | Observe records; Govern executes ALLOW/WARN and resolves or suspends other outcomes |
| Budget | Validated manual accounting plus automatic Judge usage; ceilings produce PAUSE |
| Human escalation and provenance | Local approval handler/prompt, ID-correlated approval decisions, and caller-created audit exports; durable lifecycle planned for 0.7 |
| Planner / Critic | Full-plan boundary, bound constraints, selected plan/history heuristics and supplied evidence flags |
| Judge | Explicit MockJudge; optional live OpenAI/Anthropic SDK providers, validated verdicts, timeouts and usage accounting |
| Narration / CLI | Manually populated narrator and `status`, `demo`, `version`, `approval-demo` commands; CLI status creates a fresh local narrator |
| Frameworks / integrations | LangGraph callable wrapper; additional adapters and optional exports planned for 0.8 |

Next: **0.7 live status/audit/Control Room →
0.8 adapters/integrations → 0.9 evaluation/authority → 1.0 stable schemas**.
See [supported scope and limitations](docs/limitations.md) for operational details.

### Installation & Usage
```bash
# Core package
pip install agentic-sidecar

# With LangGraph adapter (v0.4 features)
pip install "agentic-sidecar[langgraph]"

# CLI usage (v0.5)
agentic-sidecar status              # Single status snapshot
agentic-sidecar status --follow     # Repeat local snapshot (Ctrl+C to stop)
agentic-sidecar status --json       # JSON output
agentic-sidecar demo                # Interactive demo
agentic-sidecar version             # Show version
agentic-sidecar approval-demo       # Try the local approval prompt
```

See [Python API (implemented)](#python-api-implemented) below and [`examples/`](examples/) 
for runnable scripts. See [ROADMAP.md](ROADMAP.md) for the full build plan.

## Contents

- [Why](#why)
- [What it is not](#what-it-is-not)
- [Architecture](#architecture)
- [Sidecar vs. Agent Harness](#sidecar-vs-agent-harness)
- [Main Agent vs. Sidecar](#main-agent-vs-sidecar)
- [How the Decision Gate evaluates a decision](#how-the-decision-gate-evaluates-a-decision)
- [Sidecar modules](#sidecar-modules)
- [Python API (implemented)](#python-api-implemented)
- [Planned Python API](#planned-python-api)
- [Operating modes](#operating-modes)
- [Long-term: an intent propagation layer](#long-term-an-intent-propagation-layer)
- [The DeepAgentLabs ecosystem](#the-deepagentlabs-ecosystem)
- [Roadmap](#roadmap)
- [License](#license)

## Why

Autonomous agents chain a lot of reasoning and tool calls between a user's
request and a real-world effect:

```text
User → Agent → Planning → Sub-Agent → MCP → Tool → API → External System
```

Three things go wrong in that chain that permission checks alone don't catch:

- **Intent drift** — "investigate why production is slow" quietly becomes
  "restart the production database." Nobody revoked authority; the agent's
  plan just wandered.
- **Technically-allowed, contextually-wrong actions** — `refund_customer()`
  is a permitted tool call. Refunding $850 against a $500 authorization is
  not a permitted *decision*, even though the tool call succeeds.
- **No live visibility** — long-running agents act for minutes at a time
  with no answer to "what is it doing right now, and can I stop it?"

`agentic-sidecar` is designed to attach to any agent framework, but the first
release proves that against one framework only — LangGraph — before claiming
the rest (see [Design Constraints](ROADMAP.md#design-constraints)
in ROADMAP.md). Either way, it answers a different question than permission
checks or observability tools do:

> **Not** "can this agent call this tool?" — **"should it, right now, given
> what the user actually asked for?"**

## What it is not

- **Not a multi-agent framework.** The Sidecar never owns or executes the
  task plan, and it never replaces a worker agent. It may independently
  critique the Main Agent's plan or propose alternatives (the Planner
  module) — but the Main Agent decides whether to act on that critique, and
  remains the one that executes.
- **Not a content-safety guardrail.** Libraries like NeMo Guardrails or
  Guardrails AI validate a single prompt/response turn against rules. The
  Sidecar evaluates a *decision* — a tool call or plan step — against intent
  and context accumulated across an entire task.
- **Not a tracing/observability tool.** That's
  [AgenticLens](https://github.com/DeepAgentLabs/agenticlens)'s job:
  "what did the agent do, and why did it fail?" — after the fact. The
  Sidecar's job is "should it continue?" — while it's still running.
- **Not just LLM-as-a-Judge.** A judge model is one possible evaluator inside
  one Sidecar module (§ [Sidecar modules](#sidecar-modules)). Most decisions
  should never reach an LLM at all — see [Operating
  modes](#operating-modes) and the cost-control design in
  [ROADMAP.md](ROADMAP.md).
- **Not an agent harness.** It doesn't run the agent loop, manage tools, or
  handle retries — that's LangGraph's, CrewAI's, or your custom loop's job.
  The Sidecar attaches to whatever harness is already running the agent; see
  [Sidecar vs. Agent Harness](#sidecar-vs-agent-harness).

## Architecture

```text
                    USER
                      │
                 User Intent
                      │
                      ▼
              ┌───────────────┐
              │  MAIN AGENT   │
              │   EXECUTOR    │
              └───────┬───────┘
                      │
              Plans / Decisions
                      │
                      ▼
            ┌─────────────────────┐
            │   AGENTIC SIDECAR   │
            │                     │
            │ Intent Guardian     │
            │ Planner             │
            │ Critic              │
            │ Judge               │
            │ Risk Evaluator      │
            │ Policy Advisor      │
            │ Decision Gate       │
            │ Status Interpreter  │
            └──────────┬──────────┘
                       │
          Advice / Approval / Challenge
                       │
                       ▼
                 MAIN AGENT
                       │
                       ▼
                  MCP / Tools
                       │
                       ▼
                External Systems
```

The Sidecar is attached to the execution lifecycle. It does not own the
user's task and is never a second worker in the plan.

## Sidecar vs. Agent Harness

An **agent harness** (LangGraph, a custom loop, OpenAI Agents SDK, Microsoft
Agent Framework, ...) is the control and execution environment: it runs the
agent loop and manages tools, state, context, retries, and lifecycle.
`agentic-sidecar` doesn't replace that — it attaches to it as a decision-time
supervision layer:

```text
Agent Harness   = control and execution loop.
Agentic Sidecar = decision-time supervision layer for that loop.
```

The harness answers *"how do I execute this workflow?"* The Sidecar answers
*"should this decision happen, given what the human originally asked for?"*

This distinction matters most once a task fans out across a delegation
chain — Agent A → Agent B → Agent C → MCP/Tool — where each hop tends to
receive only the sub-task it needs to perform, not the original constraints
and authority behind it:

```text
Human Intent → Agent A → delegates → Agent B → delegates → Agent C → MCP/Tool
```

The Sidecar's Intent Envelope (§ [Sidecar modules](#sidecar-modules)) is
designed to travel with the task through that chain instead of being
reconstructed from conversation history at every hop — see [Long-term: an
intent propagation layer](#long-term-an-intent-propagation-layer).

Where it pays for itself: without a supervision layer, a harness typically
discovers a bad decision only after acting on it —

```text
Plan → Act → Fail → Recover → Replan
```

— versus catching it before execution:

```text
Plan → Sidecar Check → Act
          ├── ALLOW
          ├── CHALLENGE
          ├── REPLAN
          ├── BLOCK
          └── ESCALATE
```

This architecture focuses on decision reliability and control. Comparative
reliability and inference-overhead measurements are planned in the roadmap.
Architecture rationale:
[concept.md § 23](concept.md#23-sidecar-vs-agent-harness).

## Main Agent vs. Sidecar

| Main Agent | Agentic Sidecar |
| --- | --- |
| Accomplishes the task | Maintains original intent |
| Reasons about the domain problem | Independently evaluates plans and decisions |
| Selects and calls tools | Challenges questionable decisions |
| Interacts with MCP servers / APIs | Evaluates risk and checks policy |
| Executes actions | Decides when human approval is required |
| Produces the final result | Reports supplied execution state and verdicts; adapters enforce configured boundaries |

## How the Decision Gate evaluates a decision

Three modules ask three genuinely different questions about the same
proposed action, and the roadmap's job is to keep them from collapsing into
one generic "policy engine":

```text
             USER INTENT
                  │
          Intent Envelope
                  │
                  ▼
             MAIN AGENT
                  │
           proposed action
                  │
                  ▼
              SIDECAR
          ┌───────┼────────┐
          │       │        │
       Policy    Risk    Intent
          │       │        │
   "Are you    "How      "Is this
   permitted   dangerous  actually what
   to do       is this    the human
   this?"      action?"   asked you to
                          accomplish?"
          │       │        │
          └───────┼────────┘
                  │
             DECISION GATE
                  │
       ┌──────────┼──────────┐
       ▼          ▼          ▼
     ALLOW      REPLAN     ESCALATE
                              │
                            HUMAN
```

Policy and Risk are largely mechanical — permission lists, thresholds,
tool-argument patterns — and v0.1 ships them with zero LLM calls (see
[ROADMAP.md](ROADMAP.md)). Intent (v0.2) is still deterministic, not
LLM-based, but is the one that gets closest to "what the user actually
meant" so far, via structured `IntentEnvelope` constraints rather than
free-form judgment — full semantic understanding is further out on the
roadmap.

The diagram shows the supervision model. The current `Decision` type supports
all seven outcomes. Govern permits ALLOW/WARN, raises `SidecarBlockedError`
for BLOCK, invokes approvals for PAUSE/ESCALATE, and raises `SidecarSuspendedError`
for unresolved outcomes, CHALLENGE and REPLAN before dispatch.
See [supported scope](docs/limitations.md) and the [roadmap](ROADMAP.md#build-order).

## Sidecar modules

The target architecture below includes partial and planned capabilities.
See the Status table above for current behavior. Users enable only what they need:

```text
Agentic Sidecar
│
├── Intent Guardian     — Intent Envelope, alignment checks, drift detection
├── Planner              — independently evaluates the agent's plan
├── Critic                — challenges a proposed decision before it executes
├── Judge                 — optional independent (and independent-model) evaluation
├── Risk Evaluator        — classifies an action's risk before deciding whether to escalate
├── Policy Advisor        — deterministic policy rules (cheapest check, runs first)
├── Decision Gate         — turns evaluations into ALLOW / WARN / CHALLENGE / REPLAN / PAUSE / BLOCK / ESCALATE
├── Budget Guardian       — cost/token ceilings per task
├── Status Interpreter    — translates raw tool/MCP traces into human-readable narration
└── Human Escalation      — pauses execution and requests approval
```

```yaml
sidecar:
  intent:
    enabled: true
    preserve_original_intent: true
  planner:
    enabled: false        # off by default — see cost design in ROADMAP.md
  critic:
    enabled: false        # off by default — see cost design in ROADMAP.md
  policy:
    enabled: true
    source: policies.yaml
  risk:
    enabled: true
    intervention_threshold: 0.80
  judge:
    enabled: false        # off by default — see cost design in ROADMAP.md
    model: independent-model
  budget:
    enabled: true
    max_cost: 2.00
  human_approval:
    enabled: true
  on_sidecar_failure: fail_closed   # or fail_open — see ROADMAP.md
```

## Python API (implemented)

A runnable constraint example using plain callables, with no model or API key:

```python
from agentic_sidecar import Sidecar, SidecarBlockedError
from agentic_sidecar.adapters.langgraph import attach
from agentic_sidecar.intent import ConstraintBinding, IntentEnvelope, IntentGuardian, Requester

def issue_refund(order_id: str, amount: float) -> str:
    """Issue a refund for an order."""
    return f"refunded {amount} for {order_id}"

envelope = IntentEnvelope(
    goal="Refund order A100 up to $500",
    requested_by=Requester(id="support_lead"),
    constraints={"maximum_refund": 500},
)
guardian = IntentGuardian(envelope, [ConstraintBinding(
    constraint="maximum_refund", tool="issue_refund", arg_name="amount", op="lte",
)])
sidecar = Sidecar(
    on_sidecar_failure="fail_closed", mode="govern",
    intent=guardian, roles=["policy", "risk", "intent_guardian"],
)
(refund,) = attach(sidecar, [issue_refund])
print(refund("A100", 120))  # refunded 120 for A100
try:
    refund("A100", 850)
except SidecarBlockedError as exc:
    print(exc.decision.status)  # BLOCK; refund function was not called
```

Use `mode="observe"` to record verdicts while permitting wrapped calls.
`evaluate(DecisionContext(...))` computes a verdict without dispatching tools.
`set_intent(guardian)` changes the active task envelope; budget reset is separate.

For full-plan review, supply the plan explicitly:

```python
from agentic_sidecar import PlanEvaluator, Sidecar

reviewer = Sidecar(on_sidecar_failure="fail_closed", planner=PlanEvaluator(), roles=["planner"])
result = reviewer.evaluate_plan(
    [{"tool_name": "read_charge"}, {"tool_name": "issue_refund"}],
    goal="Explain the disputed charge",
)
print(result.status)  # REPLAN
```

A `before_tool_call` hook replaces the entire built-in pipeline, including
budget and evaluators. See [API reference](docs/api-reference.md),
[configuration](docs/configuration.md), and [runnable examples](examples/README.md).

## Planned Python API

The roadmap adds live status/audit/operator interfaces in 0.7 and framework adapters in 0.8.
Today `attach(sidecar, tools)` is adapter-specific and `evaluate()` accepts
`DecisionContext`. Generic whole-agent attachment will be considered when
multiple integrations demonstrate a useful shared boundary.

## Operating modes

| Mode | Current behavior |
| --- | --- |
| **Observe** | Computes/logs decisions; wrapped calls execute regardless of verdict |
| **Govern** | ALLOW/WARN execute; BLOCK stops; PAUSE/ESCALATE invoke approvals or suspend; CHALLENGE/REPLAN suspend |

Catch `SidecarSuspendedError` to return control to your agent loop. Configure
`escalation_handler=CLIApprovalHandler()` or register `on_escalation_required`
for local approval. Approve-once applies to the current call; modify-intent
reevaluates with a complete envelope; stop persists on that Sidecar instance.
Live operator controls follow in 0.7. See [0.6.1 usage and validation](docs/governance.md).

## Long-term: an intent propagation layer

The nearer-term stages above focus on runtime supervision. The
`IntentEnvelope` (§ [Sidecar modules](#sidecar-modules)) is designed to
survive being handed off — not just checked once and discarded:

```text
Human
  │
  └── Intent Envelope #182
            │
            ▼
         Agent A
            │
        delegates
            ▼
         Agent B
            │
           MCP
            ▼
         Agent C
            │
           Tool
```

If every hop in a delegation chain can answer *what was originally
requested, who authorized it, what constraints apply, what authority was
actually delegated, and whether intent has since changed* — that's no
longer just a feature of one package. It's closer to an interoperability
concern for autonomous systems generally, which is why v1.0 targets
publishing the envelope as a versioned schema in `ai-operations-spec`
rather than keeping it as an internal Sidecar structure (see the v1.0 entry
in [ROADMAP.md](ROADMAP.md#build-order)).
This is explicitly a v1.0 target, not something v0.1 needs to anticipate —
noted here because it's the reason the envelope's shape deserves care early,
even though nothing consumes it across process boundaries yet.

## The DeepAgentLabs ecosystem

```text
                 DeepAgentLabs
              Autonomous AI Systems
                      │
        ┌─────────────┼─────────────┐
        │             │             │
     OBSERVE         GUIDE        CONNECT
        │             │             │
 AgenticLens    Agentic Sidecar  Agentic MCP
        │             │             │
        └─────────────┼─────────────┘
                      │
                     TEST
                      │
                Agentic Chaos
```

| Project | Question it answers |
| --- | --- |
| [AgenticLens](https://github.com/DeepAgentLabs/agenticlens) | What did the agent do, and what happened? |
| [Agentic Chaos](https://github.com/DeepAgentLabs/agentic-chaos) | How does the agent behave when things go wrong? |
| [Agentic MCP](https://github.com/DeepAgentLabs/mcp-server) | How does the agent interact with tools and external capabilities? |
| [AgenticOps Control Tower](https://github.com/DeepAgentLabs/agenticops-control-tower) | What is deployed, what governance posture exists across agents, and how do operators manage it centrally? |
| **Agentic Sidecar** | Should the agent continue with this decision, and is it still acting according to intent? |

Each project is independently installable; none requires another as a hard
dependency (see [Cross-Project Dependencies](ROADMAP.md#cross-project-dependencies)
in the roadmap).

## Roadmap

Canonical release history, capability status, completion checklists, and future sequencing:
[ROADMAP.md](ROADMAP.md).

Original architecture proposal: [concept.md](concept.md).

## License

MIT — see [`LICENSE`](LICENSE).

## Control Tower integration boundary

AgenticOps Control Tower owns deployment registration, heartbeat, fleet
inventory and operator workflows. AgenticLens owns observability and analysis;
Agentic Evals owns scoring and release-gate computation; Agentic Sidecar owns
decision supervision (SUPERVISE); Agentic Chaos owns fault injection and
resilience experiments. AI Operations Specification owns shared semantics and
is currently draft.

Tower's native artifact readers consume existing producer outputs using an
explicit deployment/evidence link. They do not change producer behavior or
convert a failed run, failed gate, blocked action or injected fault into fleet
health. This link is a Tower-local contract, not a normative AIOS schema.
See [Control Tower's evidence contract](https://github.com/DeepAgentLabs/agenticops-control-tower/blob/main/docs/ecosystem-alignment.md).
Remote collection and operator posture views remain planned; these local
readers do not establish end-to-end integration or stable AIOS conformance.
