brain-crud

First-party

Reference: full SKILL.md

Below is the complete skill definition this hub loads when the skill is triggered — what the agent sees as its instructions, verbatim and unabridged.

# Brain CRUD

Full CRUD interface to the AXE Brain vault. Any model or agent on the fleet can use this.

## API

Base: `https://brain.axe.onl`
Auth: `X-AXE-Key` header or `Authorization: Bearer <key>`
Key: env `AXE_FLEET_KEY`

## Operations

### Save
```bash
curl -s -X POST https://brain.axe.onl/api/save \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"title":"<title>","content":"<body>","source":"<model>@<node>","context":"<type>","tags":"<comma-sep>"}'
```
Context types: `decision`, `reference`, `finding`, `note`, `status`
Source format: `model@node` (e.g. `claude@jla`, `qwen@forge`)

### Search (keyword)
```bash
curl -s "https://brain.axe.onl/api/search?q=<query>&limit=10" -H "X-AXE-Key: $AXE_FLEET_KEY"
```

### Search (semantic)
```bash
curl -s -X POST https://brain.axe.onl/api/semantic \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<text>","limit":5}'
```

### Browse recent
```bash
curl -s "https://brain.axe.onl/api/audit?limit=15" -H "X-AXE-Key: $AXE_FLEET_KEY"
```

### Neural process (full 6-layer pipeline)
```bash
curl -s -X POST https://brain.axe.onl/api/neural/process \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<text>","source":"<model>@<node>","save":false}'
```

### DPO feedback
```bash
curl -s -X POST https://brain.axe.onl/api/neural/feedback \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<prompt>","chosen":"<good>","rejected":"<bad>"}'
```

### Export training pairs
```bash
curl -s "https://brain.axe.onl/api/neural/training" -H "X-AXE-Key: $AXE_FLEET_KEY"
```

## Python (stdlib only)
```python
import urllib.request, json, os

KEY = os.getenv("AXE_FLEET_KEY")
BASE = "https://brain.axe.onl"
HEADERS = {"X-AXE-Key": KEY, "Content-Type": "application/json"}

def brain_save(title, content, source, context="note", tags="axe"):
    data = json.dumps({"title": title, "content": content, "source": source, "context": context, "tags": tags}).encode()
    req = urllib.request.Request(f"{BASE}/api/save", data=data, headers=HEADERS, method="POST")
    return json.loads(urllib.request.urlopen(req).read())

def brain_search(query, limit=10):
    req = urllib.request.Request(f"{BASE}/api/search?q={query}&limit={limit}", headers=HEADERS)
    return json.loads(urllib.request.urlopen(req).read())

def brain_semantic(query, limit=5):
    data = json.dumps({"query": query, "limit": limit}).encode()
    req = urllib.request.Request(f"{BASE}/api/semantic", data=data, headers=HEADERS, method="POST")
    return json.loads(urllib.request.urlopen(req).read())
```

Brain CRUD

Full CRUD interface to the AXE Brain vault. Any model or agent on the fleet can use this.

API

Base: https://brain.axe.onl

Auth: X-AXE-Key header or Authorization: Bearer <key>

Key: env AXE_FLEET_KEY

Operations

Save

curl -s -X POST https://brain.axe.onl/api/save \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"title":"<title>","content":"<body>","source":"<model>@<node>","context":"<type>","tags":"<comma-sep>"}'

Context types: decision, reference, finding, note, status

Source format: model@node (e.g. claude@jla, qwen@forge)

Search (keyword)

curl -s "https://brain.axe.onl/api/search?q=<query>&limit=10" -H "X-AXE-Key: $AXE_FLEET_KEY"

Search (semantic)

curl -s -X POST https://brain.axe.onl/api/semantic \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<text>","limit":5}'

Browse recent

curl -s "https://brain.axe.onl/api/audit?limit=15" -H "X-AXE-Key: $AXE_FLEET_KEY"

Neural process (full 6-layer pipeline)

curl -s -X POST https://brain.axe.onl/api/neural/process \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<text>","source":"<model>@<node>","save":false}'

DPO feedback

curl -s -X POST https://brain.axe.onl/api/neural/feedback \
  -H "X-AXE-Key: $AXE_FLEET_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"<prompt>","chosen":"<good>","rejected":"<bad>"}'

Export training pairs

curl -s "https://brain.axe.onl/api/neural/training" -H "X-AXE-Key: $AXE_FLEET_KEY"

Python (stdlib only)

import urllib.request, json, os

KEY = os.getenv("AXE_FLEET_KEY")
BASE = "https://brain.axe.onl"
HEADERS = {"X-AXE-Key": KEY, "Content-Type": "application/json"}

def brain_save(title, content, source, context="note", tags="axe"):
    data = json.dumps({"title": title, "content": content, "source": source, "context": context, "tags": tags}).encode()
    req = urllib.request.Request(f"{BASE}/api/save", data=data, headers=HEADERS, method="POST")
    return json.loads(urllib.request.urlopen(req).read())

def brain_search(query, limit=10):
    req = urllib.request.Request(f"{BASE}/api/search?q={query}&limit={limit}", headers=HEADERS)
    return json.loads(urllib.request.urlopen(req).read())

def brain_semantic(query, limit=5):
    data = json.dumps({"query": query, "limit": limit}).encode()
    req = urllib.request.Request(f"{BASE}/api/semantic", data=data, headers=HEADERS, method="POST")
    return json.loads(urllib.request.urlopen(req).read())

Metadata

Category: Agent Ops
Tier: community
Version: 1.0.0
License: MIT
Path: skills/brain-crud/SKILL.md

Use with an agent

Fetch this skill’s definition over the open API — no key required.

curl -s /v1/skills/brain-crud

View source ↗