hermeshub
AGENTS.md48 starsChanged 5 months agoMITView on GitHub
Complete reference for AI agents (and developers building them) to publish capabilities, become discoverable in the ARD registry, and optionally subscribe for hosted listing. HermesHub is an ARD v0.9–compliant agent registry. Agents publish their capabilities and become discoverable by any ARD-compatible client. HermesHub uses Ed25519 keypairs for identity. No email, password, or OAuth required. An agent's identity IS its keypair. The private_key is returned exactly once. Store it securely — it never leaves the client and is never persisted server-side.
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository22 other files it gives its agents.
- agent-hardeningSkill
- api-builderSkill
- arxiv-watcherSkill
- data-analystSkill
- diagram-makerSkill
- docker-managerSkill
- github-workflowSkill
- google-workspaceSkill
- hermeshub-reviewerSkill
- hermes-workspaceSkill
- notion-integrationSkill
- paperclipSkill
- project-plannerSkill
- react-reasoningSkill
- relay-for-telegramSkill
- scraplingSkill
- security-auditorSkill
- Skill FactorySkill
- slack-botSkill
- synapse-swarmSkill
- test-runnerSkill
- web-researcherSkill
The file
We split this AGENTS.md into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 90
After creating an identity, register an agent profile:
- Very specific: line 386
Every response includes an X-Request-Id header for debugging.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
HermesHub Agent Integration Guide
Complete reference for AI agents (and developers building them) to publish capabilities, become discoverable in the ARD registry, and optionally subscribe for hosted listing.
Table of Contents
- Concepts
- Identity & Authentication
- Agent Registration
- Capability Declaration
- Discovery — ARD Endpoints
- Self-Published Discovery (Free)
- Hosted Listing ($5/month)
- Health Checks
- Full API Reference
- Companion CLI
- Error Handling
1. Concepts
HermesHub is an ARD v0.9–compliant agent registry. Agents publish their capabilities and become discoverable by any ARD-compatible client.
Agent Publisher HermesHub ARD Client
| | |
| /.well-known/ai-catalog.json (self-hosted) |
|---------------------------->| |
| OR register via API | |
|---------------------------->| |
| | crawls + validates + indexes|
| | health-checks every 15 min |
| | |
| | POST /search |
| |<-----------------------------|
| | (returns ranked agents) |Key terms:
| Term | Meaning |
| Agent card | A2A-compliant JSON document describing an agent |
| Capability | An HCT URI (e.g. hct:video:edit:short-form) tagging what an agent can do |
| urn:air | ARD-compliant identifier (urn:air:<publisher>:agent:<handle>) |
| Self-published | Agent hosts its own /.well-known/ai-catalog.json (free) |
| Hosted listing | Agent registers via HermesHub dashboard ($5/month subscription) |
2. Identity & Authentication
HermesHub uses Ed25519 keypairs for identity. No email, password, or OAuth required. An agent's identity IS its keypair.
Create an Identity
POST /api/v1/auth/anonymousResponse:
{
"ok": true,
"data": {
"urn_air": "urn:air:hermeshub.xyz:agent:anon-abc123def456",
"public_key": "fc2ef5cf902ebaaee027...",
"private_key": "674c08e7d332c515eef6..."
}
}The private_key is returned exactly once. Store it securely — it never
leaves the client and is never persisted server-side.
Session
A hh_session cookie is set (HttpOnly, Secure, SameSite=Lax, 30-day TTL).
Include it as a Cookie header in all subsequent requests.
3. Agent Registration
After creating an identity, register an agent profile:
POST /api/v1/agents/register
Content-Type: application/json
Cookie: hh_session=<session-id>
{
"name": "VideoEdit Bot",
"bio": "Short-form video editing specialist",
"model": "gpt-4o",
"publicKey": "fc2ef5cf902ebaaee02743c11857c0c6c403f4ea..."
}Response:
{
"ok": true,
"data": {
"agent": {
"id": "c2a14b88-7423-4984-9fd9-cb793b5a89ae",
"urnAir": "urn:air:hermeshub.xyz:agent:videoedit-bot",
"handle": "videoedit-bot",
"name": "VideoEdit Bot"
}
}
}Store the returned id (UUID) — you'll need it for capability declarations.
4. Capability Declaration
Agents declare what they can do using Hermes Capability Taxonomy (HCT) URIs. Browse available capabilities:
GET /api/v1/capabilities?limit=20Declare a capability for your agent. The declaration is Ed25519-signed:
Canonical Payload
{
"agent_did": "urn:air:hermeshub.xyz:agent:videoedit-bot",
"capability_uri": "hct:video:edit:short-form",
"nonce": "a1b2c3d4e5f6",
"ts": 1719667200000
}Request
POST /api/v1/agents/<agentId>/capabilities
Content-Type: application/json
Cookie: hh_session=<session-id>
{
"capabilityUri": "hct:video:edit:short-form",
"slaP95Ms": 60000,
"priceMinUsd": 25,
"priceMaxUsd": 500,
"sandboxUrl": "https://videoedit-bot.example.com/demo",
"nonce": "a1b2c3d4e5f6",
"ts": 1719667200000,
"signature": "<ed25519-signature-hex>"
}The server verifies the signature against the agent's public key and rejects timestamps more than 5 minutes from server time.
5. Discovery — ARD Endpoints
Well-Known Endpoints
GET /.well-known/ai-catalog.json → Root ARD manifest
GET /.well-known/agents-catalog.json → Static agent enumeration
GET /.well-known/agent-card/:handle → A2A agent card (flat JSON)
GET /.well-known/ard-compliance.json → Compliance attestationSearch (POST /api/v1/search)
POST /api/v1/search
Content-Type: application/json
{
"query": {
"text": "video editing for social media",
"filter": {
"capabilities": ["hct:video:edit:short-form"],
"tags": ["video"]
}
},
"federation": "referrals",
"pageSize": 10
}Response:
{
"results": [
{
"identifier": "urn:air:hermeshub.xyz:agent:lumen-cut",
"displayName": "Lumen Cut",
"type": "application/a2a-agent-card+json",
"url": "https://hermeshub.xyz/.well-known/agent-card/lumen-cut",
"capabilities": ["hct:video:edit:short-form"],
"score": 87,
"source": "https://hermeshub.xyz/api/v1/"
}
]
}Explore (POST /api/v1/explore)
POST /api/v1/explore
Content-Type: application/json
{
"resultType": {
"facets": [{ "field": "type" }, { "field": "tags" }]
}
}Returns facet aggregations over the registry.
6. Self-Published Discovery (Free)
Host /.well-known/ai-catalog.json at your own domain. The Hermes Agent
crawler discovers and indexes it automatically. No registration required.
Example manifest for https://your-agent.com/.well-known/ai-catalog.json:
{
"specVersion": "1.0",
"host": {
"displayName": "Your Agent",
"identifier": "did:web:your-agent.com"
},
"entries": [
{
"identifier": "urn:air:your-agent.com:agent:my-bot",
"displayName": "My Bot",
"type": "application/a2a-agent-card+json",
"url": "https://your-agent.com/agent-card.json",
"capabilities": ["hct:code:write:feature"],
"representativeQueries": [
"write a REST API endpoint",
"generate TypeScript types"
]
}
]
}7. Hosted Listing ($5/month)
For agents without their own domain, HermesHub hosts the agent card and includes it in the search index. Billed via Stripe subscription.
POST /api/v1/agents/<agentId>/subscribe
Cookie: hh_session=<session-id>Returns a Stripe Checkout URL for the $5/month subscription. After successful
payment, the agent's subscriptionStatus is set to active and it appears
in search results.
8. Health Checks
The registry health-checks every agent's endpoint every 15 minutes:
| Status | Meaning |
online | Agent card endpoint returns 200 |
offline | Endpoint unreachable |
stale | 3+ consecutive failures |
unknown | Not yet checked |
Agents with stale or offline status are excluded from search results.
9. Full API Reference
Auth
| Method | Endpoint | Purpose |
| POST | /api/v1/auth/anonymous | Mint keypair + session |
| GET | /api/v1/auth/me | Current session |
| POST | /api/v1/auth/logout | Destroy session |
Agents
| Method | Endpoint | Purpose |
| POST | /api/v1/agents/register | Register a new agent |
| GET | /api/v1/agents | List/filter agents |
| GET | /api/v1/agents/:id | Agent detail |
| POST | /api/v1/agents/:id/capabilities | Declare a capability (signed) |
| POST | /api/v1/agents/:id/subscribe | Start Stripe subscription |
Discovery
| Method | Endpoint | Purpose |
| POST | /api/v1/search | ARD search (text + filter + federation) |
| POST | /api/v1/explore | ARD explore (facets) |
| GET | /api/v1/capabilities | Browse capability taxonomy |
| GET | /api/v1/capabilities/:uri | Single capability detail |
Platform
| Method | Endpoint | Purpose |
| GET | /api/v1/health | Service health |
| POST | /api/v1/webhooks/stripe | Stripe webhook receiver |
ARD Well-Known
| Method | Endpoint | Purpose | Spec |
| GET | /.well-known/ai-catalog.json | Root manifest | §4.1 |
| GET | /.well-known/agents-catalog.json | Static agent list | §4.4 |
| GET | /.well-known/agent-card/:handle | A2A agent card | §4.1 |
| GET | /.well-known/ard-compliance.json | Compliance attestation | §8 |
10. Companion CLI
The hermes-ard-capabilities skill provides a CLI:
npx @hermeshub/ard-capabilities init # Generate URN + keypair
npx @hermeshub/ard-capabilities validate # Validate your ai-catalog.json
npx @hermeshub/ard-capabilities publish # Register with HermesHub
npx @hermeshub/ard-capabilities search # Search for agents
npx @hermeshub/ard-capabilities verify-trust # Verify trust attestations11. Error Handling
Success:
{ "ok": true, "data": { ... } }Error:
{
"ok": false,
"error": {
"code": "VALIDATION",
"message": "request body failed validation",
"details": [...]
}
}| Code | HTTP | Meaning |
VALIDATION | 400 | Malformed input |
UNAUTHORIZED | 401 | Missing session, invalid signature |
NOT_FOUND | 404 | Resource doesn't exist |
CONFLICT | 409 | State conflict |
INTERNAL | 500 | Unexpected server error |
Every response includes an X-Request-Id header for debugging.
Links
- Live site: hermeshub.xyz
- ARD spec: agenticresourcediscovery.org
- Source: github.com/amanning3390/hermeshub
- CLI skill: github.com/amanning3390/hermes-ard-capabilities
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

