ACI v1.0 requires three independent implementations from teams or individuals who aren't the spec steward. That's the hard gate. Help get us there.
How to participate
- Fork the Level 1 template or implement directly from SPEC.md
- Deploy your manifests publicly
- Validate with
python3 validator/validate.py https://your-site.com
- Comment below with:
- Organization or project name
- Implementation URL
- Conformance level (1, 2, or 3)
- Validator output
- What worked, what didn't, what was ambiguous
Progress
| # |
Organization |
URL |
Level |
Status |
| 1 |
Empire Labs (steward) |
https://empirelabs.com.au |
3 (candidate) |
Validated at 100/100 |
| 2 |
— OPEN — |
— |
— |
Waiting for you |
| 3 |
— OPEN — |
— |
— |
Waiting for you |
Goal: Three implementations across at least two continents, with at least one non-trivial Level 2 or Level 3 deployment.
Why participate?
- Shape a specification that's still in draft — your feedback goes directly into v1.0
- Get listed as a founding implementer
- If you're building autonomous systems, you need this problem solved anyway
Read the full story
We published a deep dive on ACI here: https://empirelabs.hashnode.dev (article going live)
Questions we specifically need critical review on
- Is the five-manifest model complete, or is something missing?
- Are the discovery mechanisms (llms.txt + /.well-known/aci + DNS TXT) sufficient for agent discovery?
- Are the conformance levels scoped appropriately?
- Are the identifier grammar rules practical for real-world use?
- What's the biggest barrier to implementing ACI at your organization?
ACI v1.0 requires three independent implementations from teams or individuals who aren't the spec steward. That's the hard gate. Help get us there.
How to participate
python3 validator/validate.py https://your-site.comProgress
Goal: Three implementations across at least two continents, with at least one non-trivial Level 2 or Level 3 deployment.
Why participate?
Read the full story
We published a deep dive on ACI here: https://empirelabs.hashnode.dev (article going live)
Questions we specifically need critical review on