Skip to content
Little Canary 0.5.1 · Python 3.9+ · Apache-2.0 published-release licenseOpen source · inspectable boundary

Experimental prompt-injection sensing

Inspect untrusted input before an agent acts.

Little Canary sends untrusted language through structural checks and a powerless sacrificial model, then reports routing and inspection coverage separately.

Verify the pinned CLIno project files required
$ uvx --from little-canary==0.5.1 little-canary --version
Release
v0.5.1
Published on PyPI
Runtime
Python 3.9+
Apache-2.0 published-release license
First run
One pinned command
No project files required
Scope
Bounded tool
Not a system-level guarantee

Why this exists

Make the evidence boundary visible.

An authority-bearing agent can receive untrusted language before it is clear whether inspection took place. Little Canary separates routing disposition from inspection coverage so a degraded or unavailable check is not mistaken for a completed clean inspection.

01

Structural preflight

Checks known input shapes before a model call is considered.

02

Powerless canary

Uses a sacrificial model with no application tools or authority.

03

Explicit coverage

Separates routing disposition from whether the enabled inspection layers ran.

The boundary

Useful evidence is not a blanket guarantee.

Little Canary is an inbound risk sensor, not a security guarantee or agent runtime. It does not claim universal detection, formal security, or aggregate accuracy.

Compare it with the other open-source tools →
Read the technical note →

Hermes Agent plugin catalog

Little Canary's native Hermes Agent plugin is listed in the Nous Research Hermes Agent plugin catalog, pinned to 0.4.0 at commit 61855a9. Hermes Labs submitted the entry and a Nous maintainer merged it.

Date. Listed 24 September 2026; updated to 0.4.0 on 2 October 2026.

Source. Catalog entry PR #120002 ↗ · 0.4.0 update PR #127925 ↗ · Catalog record ↗

Evidence class. Hermes-submitted external acceptance

Does not prove. It does not show Nous Research endorsement, how many Hermes Agent users run the plugin, or detection efficacy. In Hermes Agent, a BLOCK verdict blocks downstream tools, but the original prompt still reaches the model.

Orbit optional security dependency

Orbit, an independent local-first observability and security tool for LangGraph agents on Ollama, declares little-canary>=0.3.0 as its optional security extra. Its prompt-injection module imports little_canary.CanaryDetector behind an optional-dependency fallback. The pinned CI workflow skips installing that extra, citing the lack of a Linux wheel for little-canary.

Source. Optional security extra declaration ↗ · CanaryDetector import and fallback ↗ · Historical CI skip explanation ↗

Evidence class. Independent optional dependency declaration and source import

Does not prove. These immutable source records do not establish verified Orbit CI execution, deployment, endorsement, or detection efficacy. The historical CI explanation does not establish current Linux wheel availability.