oasyce
v3.2.0Oasyce Protocol — decentralized AI capability marketplace. Register data assets with manual pricing, list AI capabilities, run autonomous agent tasks, trade shares on bonding curves, and operate your node on the network. Supports scheduled scan-register-trade cycles, capability delivery with escrow...
Installation
Oasyce Protocol Skill
Decentralized AI capability marketplace — data rights + AI capabilities + autonomous agent + oracle feeds + agent identity + P2P node.
Prerequisites
pip install oasyce # core protocol (required)
pip install datavault # local data scanner (optional)
oasyce doctor # verify everything is ready
Agent Scheduler (Autonomous Mode)
Install the plugin, configure once, let it earn OAS on its own.
The agent scheduler runs periodic cycles: scan → classify → auto-register → auto-trade based on your trust level settings.
Commands
oasyce agent start # enable scheduler (persists)
oasyce agent stop # disable scheduler
oasyce agent status # show status, last run, next run, stats
oasyce agent run # trigger one immediate cycle
oasyce agent config # show current config
oasyce agent config --interval 12 # run every 12 hours (default: 24)
oasyce agent config --scan-paths ~/data,~/models # directories to scan
oasyce agent config --auto-trade # enable auto-buying capabilities
oasyce agent config --no-auto-trade # disable auto-buying (default)
oasyce agent config --trade-tags nlp,vision # only buy capabilities matching tags
oasyce agent config --trade-max-spend 20.0 # max OAS per trade cycle
How It Works
- Scan: scans configured directories for registerable assets
- Classify: evaluates sensitivity (public/internal/sensitive), assigns confidence
- Register: auto-approves based on trust level (manual / semi-auto / full-auto)
- Trade: if enabled, searches for capabilities matching trade tags and buys within budget
All runs are logged to SQLite. View history via Dashboard (Automation page) or oasyce agent status --json.
Trust Levels
| Level | Behavior |
|---|---|
| Manual (0) | Every action needs your approval |
| Semi-Auto (1) | Low-value + public sensitivity auto-approved |
| Full-Auto (2) | Everything auto-approved, notifies on anomalies |
Data Assets
Register with Pricing Control
oasyce register <file> --owner <NAME> --tags tag1,tag2
# Pricing strategies
oasyce register data.csv --price-model auto # bonding curve (default)
oasyce register data.csv --price-model fixed --price 5.0 # fixed price: 5 OAS
oasyce register data.csv --price-model floor --price 2.0 # bonding curve, min 2 OAS
# Rights declaration
oasyce register data.csv --rights-type original # 1.0x multiplier (default)
oasyce register data.csv --rights-type co_creation
--co-creators '[{"address":"A","share":60},{"address":"B","share":40}]'
oasyce register data.csv --rights-type licensed # 0.7x multiplier
oasyce register data.csv --rights-type collection # 0.3x multiplier
oasyce register data.csv --free # attribution only, no pricing
Pricing Models:
- auto — Bonding Curve: price rises with demand, factors in scarcity, quality, freshness
- fixed — Creator sets exact price, buyers pay this amount
- floor — Bonding Curve with minimum: market pricing, but never below your floor
Search & Quote
oasyce search <tag> # search by tag/keyword
oasyce quote <asset_id> # bonding curve spot price
oasyce price <asset_id> # detailed pricing factors
oasyce price-factors <asset_id> # factor breakdown
Trade
oasyce buy <asset_id> --buyer <ID> --amount 10.0
oasyce shares <owner_id> # check holdings
Asset Info
oasyce asset-info <asset_id> # full 5-layer breakdown (OAS-DAS)
oasyce asset-validate <asset_id> # validate against standard
AI Capability Marketplace
List your AI capability on the market. Others discover and invoke it. Settlement via escrow: lock → call → settle (5% protocol fee).
Register (List) a Capability
oasyce capability register --name "Translation API"
--endpoint https://api.example.com/translate
--api-key sk-xxx --price 0.5 --tags nlp,translation
Browse & Invoke
oasyce capability list [--tag nlp] [--provider addr]
oasyce capability invoke CAP_ID --input '{"text":"hello"}'
oasyce capability earnings --provider addr # provider earnings
oasyce capability earnings --consumer addr # consumer spending
Discovery (4-Layer Recall→Rank)
oasyce discover --intents "翻译" --tags nlp # broad recall + trust-ranked results
Dashboard
oasyce start # opens http://localhost:8420
Navigate to Market tab to browse, invoke, and trade capabilities.
Deep links: http://localhost:8420/#explore/CAP_ID to directly view a capability.
Oracle Feeds
Oracle feeds bridge real-world data into the network with economic guarantees.
Feed Types
| Type | Risk Factor | Recommended Bond |
|---|---|---|
| weather | 0.5× | 50 OAS |
| price | 3.0× | 300 OAS |
| time | 0.3× | 30 OAS |
| event | 2.0× | 200 OAS |
| sensor | 1.5× | 150 OAS |
Programmatic
from oasyce_core.oracle import OracleRegistry
from oasyce_core.oracle.feeds import WeatherFeed, TimeFeed
registry = OracleRegistry(provider_id="my_node")
registry.register_feed(WeatherFeed())
result = registry.execute("weather", {"location": "Shanghai"})
Agent Identity
Every participant carries portable reputation across all asset types.
Trust Tiers
| Tier | Min Rep | Max Access | Time to Reach |
|---|---|---|---|
| sandbox | R < 20 | L0 | New (default) |
| basic | R ≥ 20 | L0-L1 | ~0.7 days |
| verified | R ≥ 50 | L0-L2 | ~2.7 days |
| trusted | R ≥ 75 | L0-L3 | ~4.3 days |
Cross-asset composite score: data access (40%) + capability invocation (35%) + oracle provision (25%).
Consensus (PoS)
oasyce consensus status # epoch, slot, validators
oasyce consensus validators [--all] # list validators
oasyce consensus register --stake 10000 # become validator
oasyce consensus delegate <validator_id> --amount 500
oasyce consensus undelegate <validator_id> --amount 200
oasyce consensus rewards [--epoch N]
oasyce consensus slashing [--validator X]
oasyce consensus exit # voluntary exit
oasyce consensus unjail # unjail after penalty
Governance
oasyce governance propose --title "..." --description "..." --changes '[...]' --deposit 1000
oasyce governance vote <proposal_id> --option yes|no|abstain
oasyce governance tally <proposal_id>
oasyce governance list [--status voting|passed|rejected]
oasyce governance params [--module consensus]
Dispute & Resolution
oasyce dispute <asset_id> --reason "..."
oasyce resolve <asset_id> --remedy delist|transfer|rights_correction|share_adjustment
P2P Node & Sync
oasyce start # Core + Dashboard (recommended)
oasyce node info # show identity (Ed25519 pubkey)
oasyce node peers # list connected peers
oasyce node ping <host:port> # ping another node
oasyce sync --status # sync status
oasyce sync --peers http://host:9528 # sync from peers
Testnet
oasyce testnet onboard # one-click: identity + faucet + register + stake
oasyce testnet faucet # claim free OAS
oasyce testnet init --validators 4 --output ./testnet
Fingerprint Watermarking
oasyce fingerprint embed <file> --caller <id>
oasyce fingerprint extract <file>
oasyce fingerprint trace <fingerprint_hex>
oasyce fingerprint list <asset_id>
Access Control (L0-L3)
oasyce access query <asset_id> --agent <id> # L0: aggregated stats
oasyce access sample <asset_id> --agent <id> --size 10 # L1: redacted sample
oasyce access compute <asset_id> --agent <id> --code "..." # L2: TEE compute
oasyce access deliver <asset_id> --agent <id> # L3: full delivery
Dashboard
oasyce start # Core + Dashboard (port 8420)
| Page | URL | Description |
|---|---|---|
| Home | #home |
Register data assets and list capabilities |
| My Data | #mydata |
Manage your assets and capability earnings |
| Market | #explore |
Browse, trade, invoke capabilities |
| Automation | #auto |
Agent scheduler, approval queue, trust rules |
| Network | #network |
Node identity, AI config, consensus, watermark |
Deep links: #explore/ASSET_ID, #network/consensus, #network/watermark
Local Data Inventory (Optional — DataVault)
datavault scan [path] # scan directory
datavault classify <file> # classify single file
datavault report [path] [--format json] # generate report
Diagnostics
oasyce doctor # health check
oasyce demo # run full pipeline demo
oasyce info # project info and links
All Commands Support --json
Every command accepts --json for programmatic output, making it easy for agents to parse results.
When to Use
- Autonomous agent operation (scheduled scan/register/trade)
- Data registration with manual pricing control (auto/fixed/floor)
- AI capability listing, discovery, invocation, settlement
- Oracle feed registration and querying
- Agent identity and reputation management
- Consensus participation, staking, governance voting
- Fingerprint watermarking and provenance verification
- Testnet onboarding and demos
When NOT to Use
- General file management (mv/cp/rm — use standard tools)
- General crypto questions unrelated to data rights
- Browser-based web3 wallet interactions