> For the complete documentation index, see [llms.txt](https://docs.lpp-minduniverse.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lpp-minduniverse.org/lingua-pactum-protocol-lpp-documentation/4.-governance-objects/4.4-canonical-intent-commitment.md).

# 4.4 Canonical Intent Commitment

The **Canonical Intent Commitment** is the normalized intent commitment against which an evolving task, delegation, or action trajectory can be evaluated.

It exists between Action Candidate formation and trajectory-level Semantic Verification.

The canonical sequence is:

```
Human-Authorized Intent
        ↓
Authority–Intent Binding
        ↓
Task / Agent Evolution
        ↓
Action Candidate
        ↓
Canonical Intent Commitment
        ↓
Semantic Verification
```

#### Purpose

AI task trajectories rarely preserve their original wording exactly.

Agents may:

* paraphrase,
* decompose,
* summarize,
* translate,
* delegate,
* optimize,
* or convert natural language into structured tool calls.

A governance system therefore cannot require literal string identity.

Instead, it requires a stable semantic commitment against which the current trajectory can be evaluated.

The Canonical Intent Commitment serves that role.

#### Canonical Does Not Mean Model-Generated Authority

The word **Canonical** refers to the normalized reference representation used for verification.

It does not mean that an AI model can redefine what the authorized intent was.

Normalization must preserve the legitimate upstream meaning.

The system must not transform:

```
Authorized Intent H
```

into:

```
Model-preferred Intent H'
```

and then treat `H'` as the original authority basis.

#### Relationship to Semantic Verification

Semantic Verification compares the evolving trajectory or proposed action against the applicable authorized intent commitment.

Possible results are later defined as:

```
MATCH
CONTESTED
MISMATCH
UNVERIFIABLE
```

The Canonical Intent Commitment provides the reference necessary for that comparison.

#### Material Mutation

If the current action materially departs from the authorized commitment, previous authorization cannot simply be inherited.

The governing principle is:

> **Material Mutation ⇒ Re-Admission**

The exact trajectory semantics and verification model are defined in Chapter 6.

#### Current Public Specification Boundary

GitBook v2.0 defines the function and position of the Canonical Intent Commitment.

The detailed public machine-readable schema is reserved for Appendix 14.2 and the corresponding canonical technical specification.

No additional normative fields are assumed here.

***
