REPOSITORY ANALYSIS

eliza

elizaOS · SDK & Decision Frameworks

An optional TypeSafe HTTP adapter in Eliza’s source, not registered with the Agent runtime by default.

E2 · source reviewedPublic material or source evidence was reviewed. A fixed-version source review requires checking a pinned revision; this is not a reproduction, performance endorsement, or safety guarantee.

Open GitHub repository ↗ Back to project radar

Stars: 19,468
Forks: 5,761
License: MIT
Last commit: 2026-09-24

JEV · DECISION POINT

What Jev does here

Only an explicit systemOne call sends state and questions, returning validated typed answers.

WHY IT HELPS

Provides a reusable server-side module; its presence is not evidence of deployed Jev adoption.

SOURCE NOTE

The adapter is not registered with the runtime by default; source presence does not establish real adoption or deployment.

REVIEWED SOURCE LINKS

Source 1 ↗Source 2 ↗Source 3 ↗

EDITORIAL ANALYSIS · TASK AND IMPLEMENTATION

What task does this project fit?

Add typed judgments to an Eliza agent without changing the active chat provider.

Material analysed：2026-10-24

From input to action

Validate JSON state and the supported question subset.
POST through a bounded HTTPS client.
Validate answer IDs, types and distributions before the caller applies policy.

Implementation details that matter

The adapter implements a documented subset

Its README supports plain instructions/criteria but excludes newer structured-instruction and score-map extensions. Native API compatibility should not be assumed for every feature.

Inspect the supporting source ↗

Errors deliberately omit provider bodies

client.ts refuses credential-bearing endpoint URLs, disables redirects, cancels non-success bodies and emits sanitized error categories without automatic retry.

Inspect the supporting source ↗

When to choose it

Use it as an independent semantic decision service in an existing agent. Authorization, tool execution and fallback remain the application's responsibility.

How to inspect it yourself

Inspect systemOne and validateResponse with malformed IDs, wrong question types and a cancellation fixture. Keep transport-contract tests separate from real model-quality evaluation.

Limits of this review

No running Eliza integration or paid inference was tested.
The README's offline tests do not establish actual service availability or decision accuracy.

Primary material used in this analysis