skywire logo

Skywire Reward Eligibility Rules¶

We have transitioned to a new system with daily reward distribution

The rules in this article may change at any time, depending on if there are problems
We will attempt to address any issues reported via @skywire Telegram channel
Reward corrections are not possible after the rewards for a day have been distributed.

The required minimum Skywire version will be incremented periodically.

Introduction¶

Updates to this article will be followed by a notification via the official Skywire PSA channel on Telegram.

All information about rewards will be published here. Please ask for clarification in the @skywire Telegram channel if some things appear to not be covered.

Please join @SkywirePSA for public service announcements (PSA) regarding the skywire network, update notices, changes to this article, etc.

Reward distribution notifications are on telegram @skywire_reward.

Information about reward distribution as well as other metrics for the skywire network can be found at theskywirenetwork.net

Reward Pools¶

816000 Skycoin are distributed annually to those visors which meet the minimum uptime and the other requirements listed below, in two equally sized reward pools.

A total of up to ~1117.808 Skycoin per pool are distributed daily in non leap-years.

A total of up to ~1114.754 Skycoin per pool are distributed daily in leap-years.

Pool 1: Presence¶

The presence pool reward for a day is evenly divided among all eligible visors on the basis of having met uptime and other requirements for the previous day, with IP and MAC address deduplication applied. All architectures are eligible for the presence pool.

Regional saturation scaling is applied to the presence pool to promote geographic diversity. See Regional Saturation Scaling for details.

Pool 2: Bandwidth¶

The bandwidth pool reward for a day is distributed based on the amount of transport bandwidth each visor sent during the previous day — sender-pays. A visor accrues bandwidth credit for the bytes it transmitted on each transport, with its per-edge sent counter capped by the counterparty's reported recv so a one-sided inflation cannot pump credit. Only visors whose sender-side total exceeds a minimum threshold are eligible for the bandwidth pool.

The pool is divided using a square-root scaling of sent bytes — analogous to how regional saturation scales the presence pool, but applied to bandwidth amount instead of country IP count:

visor_weight    = sent_bytes ^ bandwidth_exponent     (default 0.5)
visor_share     = visor_weight / sum(all visor_weights)
visor_reward    = visor_share * pool_2_budget

With the default exponent of 0.5, a 100× bandwidth lead over another sender produces only a ~10× share advantage. This prevents any single high-throughput visor from dominating the pool while still rewarding heavier senders more than lighter ones. Setting the exponent to 1.0 reverts to strict bytes-proportional distribution; 0 weights every qualifying sender equally.

Each transport edge is evaluated independently: a visor earns bandwidth credit for what it sent regardless of whether the counterparty is also reward-eligible. Credit only flows to the sender, so an ineligible peer was never going to receive a share — its eligibility is not checked.

Visors which do not meet the minimum bandwidth threshold will still receive rewards from the presence pool, but will not receive any share of the bandwidth pool. In practice the effective floor is also bounded by Skycoin's 6-decimal precision: any visor whose share of the day's bandwidth pool would round to less than 0.000001 SKY receives nothing for the day, even if its sent bytes cleared the configured minimum threshold.

The bandwidth considered for this pool excludes same-LAN transports (where both edges share the same external IP address).

Rules & Requirements¶

To receive Skycoin rewards for running skywire, the following requirements must be met:

1) Minimum skywire version v1.3.50 - Cutoff May 22nd 2026
2) 75% uptime per day minimum is required to be eligible to receive rewards
3) Visors must be running on the skywire production deployment with a config that is updated on every version. No default keys or addresses of this configuration may be removed - but you can add keys where applicable.
4) Only 1 (one) visor per machine - virtual machines are ineligible
5) Up to 8 (eight) visors may each receive 1 (one) reward share per location (ip address) - with 8 shares divided between all visors at that ip address which achieved the minimum uptime. (Temporary exception: for the month of June 2026 only, the per-IP cap is raised to 12 shares — see Per IP Limit for the rationale; the cap auto-reverts to 8 on 2026-07-01.)
6) A valid skycoin address must be set for the visor
7) The visor must be connected to the DMSG network
8) Transports can be established to the visor
9) The visor has at least 2 Transports observed at some point during the same period as uptime (checked hourly)
10) The visor responds to Transport Setup-Node requests
11) The visor can have routes established to it & responds to pings over routes
12) The visor reports transport bandwidth to the transport discovery on transport re-registration - the basis for bandwidth-based rewards
13) The visor produces a survey when queried over dmsg by any keys in the survey_whitelist array by default
14) Regional saturation scaling is applied to the presence pool to incentivize geographic diversity of the network

Exceptions for Deployment Changes with dmsghttp-config (Chinese users)¶

All the production deployment services may be accessed by the visor over the dmsg network when the visor runs with a dmsghttp config.

This type of config is generated automatically based on region via:

skywire cli config gen -b --bestproto

to circumvent ISP blocking of http requests.

In order to bootstrap the visor's to connection to the dmsg network (via TCP connection to an individual dmsg server) the dmsghttp-config.json is provided with the skywire binary release.

In the instance that the skywire production deployment changes - specifically the dmsg servers: * it will be necessary to update to the next version or package release which fixes the dmsg servers. OR * it will be necessary to manually update the dmsghttp-config.json which is provided by your skywire installation.

Currently, there is no mechanism for updating the dmsghttp-config.json which does not require an http request ; a request which may be blocked depending on region.

In this instance, the visor will not connect to any service because it is not connected to the dmsg network, so it will not be possible for the visor to accumulate uptime or for the reward system to collect the survey, which are prerequisites for reward eligibility.

As a consequence of this; any visors running a dmsghttp-config, and hence any visors running in regions such as China, the minimum version requirement for obtaining rewards is not only the latest available version, but the latest release of the package unless the dmsghttp-config.json is updated manually within your installation.

Verifying Requirements & Eligibility¶

1) Version¶

View the version of skywire you are running with:

skywire cli -v
skywire visor -v

NOTE: the uptime tracker and other services consider the visor's config version - not the actual binary version.

It is expected that the visor's config is re-generated on every update in order to include any config changes which may have been introduced

The update deadlines specify the version of software required as of (i.e. on or before) the specified date in order to maintain reward eligibility:

Reward eligibility after 02-01-2026 requires Skywire v1.3.33

Requirement established 01-04-2026

Rewards Cutoff date for updating 02-01-2026

Reward eligibility after 02-25-2026 requires Skywire v1.3.34

Requirement established 02-10-2026

Rewards Cutoff date for updating 02-25-2026

Reward eligibility after 03-20-2026 requires Skywire v1.3.36

Requirement established 03-06-2026

Rewards Cutoff date for updating 03-20-2026

Reward eligibility after 04-18-2026 requires Skywire v1.3.40

Requirement established 04-04-2026

Rewards Cutoff date for updating 04-18-2026

Reward eligibility after 04-19-2026 requires Skywire v1.3.41

Requirement established 04-05-2026

Rewards Cutoff date for updating 04-19-2026

Reward eligibility after 04-23-2026 requires Skywire v1.3.42

Requirement established 04-09-2026

Rewards Cutoff date for updating 04-23-2026

Reward eligibility after 04-23-2026 requires Skywire v1.3.43

Requirement established 04-09-2026

Rewards Cutoff date for updating 04-23-2026

Reward eligibility after 05-09-2026 requires Skywire v1.3.47

Requirement established 04-25-2026

Rewards Cutoff date for updating 05-09-2026

Reward eligibility after 05-22-2026 requires Skywire v1.3.50

Requirement established 05-08-2026

Rewards Cutoff date for updating 05-22-2026

Reward eligibility after 06-10-2026 requires Skywire v1.3.59

Requirement established 05-27-2026

Rewards Cutoff date for updating 06-10-2026

Uptime¶

Daily uptime statistics for all visors may be accessed via the

uptime tracker

or using skywire cli

skywire cli ut -n0 -k <public-key>

For example with a locally running visor:

skywire cli ut -n0 -k $(skywire cli visor pk)

The uptime for all visors are tracked via the visor connecting to the uptime tracker at regular intervals (currently, every 5 minutes).

It's possible to check uptime for a given visor's public key in the following ways:

Directly in the browser

Using an http client such as curl with an appropriate json parser such as jq - for exaple:

$ curl -sL https://ut.skywire.skycoin.com/uptimes?v=v2 | jq '[.[] | select(.pk == "02006214d489aee383dc14dea22d1c308a547520f545b914c1e476a7a87064e77b" or .pk == "02001319d25a66609b64cc389bd0dcee6d5b4ab93bd139186c47d1d6cbe955e7e1")]'
[
  {
    "pk": "02001319d25a66609b64cc389bd0dcee6d5b4ab93bd139186c47d1d6cbe955e7e1",
    "on": false,
    "version": "v1.3.30",
    "daily": {
      "2025-01-03": "86.11",
      "2025-01-05": "88.89",
      "2025-01-06": "100.00",
      "2025-01-07": "100.00",
      "2025-01-08": "87.15"
    }
  },
  {
    "pk": "02006214d489aee383dc14dea22d1c308a547520f545b914c1e476a7a87064e77b",
    "on": true,
    "version": "v1.3.30",
    "daily": {
      "2025-01-03": "100.00",
      "2025-01-04": "100.00",
      "2025-01-05": "100.00",
      "2025-01-06": "100.00",
      "2025-01-07": "95.14",
      "2025-01-08": "99.65",
      "2025-01-09": "71.18"
    }
  }
]

Using skywire cli ut is the recommended approach, as it avoids rate-limiting of requests by caching the data which is fetched from the uptime tracker

$ skywire cli ut --help
query uptime tracker

Check local visor daily uptime percent with:
 skywire-cli ut -k $(skywire cli visor pk)
Set cache file location to "" to avoid using cache files



Flags:
  -m, --cfa int      update cache files if older than n minutes (default 5)
      --cfu string   UT cache file location. (default "/tmp/ut.json")
  -n, --min int      list visors meeting minimum uptime (default 75)
  -o, --on           list currently online visors
  -k, --pk string    check uptime for the specified key
  -s, --stats        count the number of results
  -t, --stats2       count of versions
  -u, --url string   specify alternative uptime tracker url (default "http://ut.skywire.skycoin.com")

Architecture¶

All computer hardware and architectures are eligible for both reward pools.

Deployment¶

The deployment your visor is running on can be verified by comparing the services configured in the visor's .json config against conf.skywire.skycoin.com

cat /opt/skywire/skywire.json

The service configuration will be automatically updated any time a config is generated or regenerated.

For those visors in China or those running a dmsghttp-config, compare the dmsghttp-config of your current installation with the dmsghttp-config on the develop branch of github.com/skycoin/skywire

The same data in a different format should be displayed in the dmsg-discovery all_servers page. Ensure that the dmsghttp-config.json in your installation has the same ip addresses and ports for the dmsg server keys.

The data from the dmsg discovery should be considered authoritative or current.

Per Machine Limit¶

Virtual machines are not eligible for rewards, as there is no way to limit the number of instances of skywire running on the same machine if they are running in virtual machines.

The visor is determined to be running on a virtual machine if a hypervisor is listed in it's survey such as kvm, xenhvm, or hyperv - (not to be confused with a skywire hypervisor)

Multiple instances of skywire which are otherwise determined to be running on the same machine (via their mac address) will have one reward share divided between all the skycoin reward addresses which are set for those visors.

Per IP Limit¶

A maximum of 8 reward shares per ip address is divided between all visors which made uptime at a given ip address.

For example, if 9 visors at a given ip address meet the minimum uptime requirement, the reward share for each visor will be 8/9ths of a share (~0.8888) which sum to a total of ~7.99999 shares.

Temporary exception — June 2026 only: the per-IP cap is raised from 8 to 12 shares for calc dates within June 2026 (UTC). Rationale: a specific operator was unable to relocate excess visors off a single IP address because OS-level update infrastructure was blocked by ongoing bandwidth-pool issues this month, and several boards in their official miner died reducing the visor count to 11+1 at one location. The cap reverts automatically to 8 on 2026-07-01 with no further action required.

Regional Saturation Scaling¶

To incentivize geographic diversity of the skywire network, a regional saturation scaling factor is applied to the presence pool. This adjusts each visor's reward share based on how many unique IP addresses are present in the visor's country.

The scaling uses a power function on the number of unique IP addresses per country:

country_weight = unique_ips_in_country ^ exponent
country_proportion = country_weight / sum(all_country_weights)

The default exponent is 0.5 (square root). Each country's total reward allocation is then divided among its visors according to their existing shares (after IP and MAC deduplication).

This means: - The first visor in a previously unrepresented country receives the highest per-visor reward - Each additional unique IP address in a country still increases that country's total allocation, but with diminishing returns - Adding more visors behind an existing IP address does not change the country's unique IP count and therefore does not increase its regional allocation - No country is excluded from rewards -- all eligible visors receive rewards regardless of location

Example: If Country A has 100 unique IPs and Country B has 1 unique IP: - Country A's weight is sqrt(100) = 10 - Country B's weight is sqrt(1) = 1 - Country A gets 10/11 (~91%) of the combined allocation, not 100/101 (~99%) - Country B's single visor is worth approximately 10x more per-IP than each of Country A's

The geographic location of each visor is determined by GeoIP lookup of the visor's IP address using the embedded MaxMind GeoLite2-City database at the country level.

The exponent can be adjusted via the --sat-exp flag (default 0.5). Setting it to 1.0 disables regional saturation scaling entirely.

Skycoin Address¶

The skycoin address to be rewarded can be set from the cli:

skywire cli reward <skycoin-address>

$ skywire cli reward --help

    reward address setting

    Sets the skycoin reward address for the visor.
    The config is written to the root of the default local directory

    this config is served via dmsghttp along with transport logs
    and the system hardware survey for automating reward distribution

Flags:
      --all   show all flags

$ skywire cli reward 2jBbGxZRGoQG1mqhPBnXnLTxK6oxsTf8os6
Reward address:
  2jBbGxZRGoQG1mqhPBnXnLTxK6oxsTf8os6

$ skywire cli reward --help

    skycoin reward address set to:
    2jBbGxZRGoQG1mqhPBnXnLTxK6oxsTf8os6

Flags:
      --all   show all flags

or via the hypervisor UI.

the example above shows the genesis address for the skycoin blockchain. Please do not use the genesis address.

It is highly recommended to set the reward address in the file '/etc/skywire.conf' by adding this line to the file:

echo "REWARDSKYADDR=('2jBbGxZRGoQG1mqhPBnXnLTxK6oxsTf8os6')" | sudo tee -a /etc/skywire.conf

PLEASE DO NOT USE THE GENESIS ADDRESS!

Add your skycoin address there and run:

skywire-autoconfig

on linux (assumes you have installed the package)

If this file does not exist for you, it can be created with

skywire cli config gen -q | sudo tee /etc/skywire.conf

If you do this, YOU MUST UNCOMMENT THE FOLLOWING LINES OF THE FILE:

PKGENV=true
BESTPROTO=TRUE

you can open the file in an editor to make that change, for instance nano

sudo nano /etc/skywire.conf

Connection To DMSG Network¶

For any given visor, the system hardware survey and transport setup-node survey are collected hourly by the reward system over dmsg.

This can be verified by examining the visor's logging:

[DMSGHTTP] 2024/10/09 - 22:31:45 | 200 |        47.2µs |                 | 03714c8bdaee0fb48f47babbc47c33e1880752b6620317c9d56b30f3b0ff58a9c3:51405 | GET      /health
[DMSGHTTP] 2024/10/09 - 22:31:46 | 200 |     193.325µs |                 | 03714c8bdaee0fb48f47babbc47c33e1880752b6620317c9d56b30f3b0ff58a9c3:51457 | GET      /node-info

Bandwidth data is reported separately by each visor to the transport discovery on transport re-registration — it is no longer collected from the visor over dmsg as a per-day CSV. See Transport Bandwidth and Metrics.

The collected surveys should be visible in the survey index here:

theskywirenetwork.net/log-collection/tree

An example of one such entry:

├─┬025e3e4e324a3ac2771e32b798ca3d8859e585ac36938b15a31d20982de6aa31fc
│ ├──health.json     Age: 13m5s {"build_info":{"version":"v1.3.21","commit":"5131943","date":"2024-04-13T15:03:26Z"},"started_at":"2024-05-07T08:52:09.895919222Z"}
│ ├──node-info.json          "v1.3.21"
│ └──tp.json         Age: 6m56s []

Note: the system survey (node-info.json) will only exist if the reward address is set.

If your visor is not generating such logging or errors are indicated, please reach out to us on telegram @skywire for assistance

Transportable¶

It is not required that a visor run any service, such as a vpn or socks5 proxy server, which permits direct access to the internet from your ip address. However, it is required that the visor is able to act as a hop along a route. A module is active at runtime which checks that transports may be established to that visor - the visor creates a dmsg transport to itself every few minutes to ensure transportability. If it's not possible to create a dmsg transport to the same visor after three attempts,the visor will shut down automatically. It is expected that the visor will be restarted by a process control mechanism if the visor shuts down for any reason. In the officially supported linux packages, systemd will restart the visor if it stops; regardless of the exit status of the process.

Transports¶

To be useful for routing traffic, a visor must have a minimum of 2 transports, observed at some point during which the uptime is being considered.

Note that this requirement depends on at least 2 public visors being online on the network at all times, and that your visor is running with public autoconnect logic enabled (which is the default).

in the instance that the visor does not have transports, it is recommended to manually create transports to the two public visors with the most transports, which can be observed with:

skywire cli pv -t

Transport Setup Node¶

The visor must respond to transport setup-node requests. The transport setup-nodes configured for the visor are included in the survey and verified as an eligibility requirement for rewards.

Routes are established through the network via setup-nodes. When a setup-node contacts your visor to establish a route, the visor must accept and process the request. Visors which do not have the correct setup-nodes configured, or which are unable to respond to setup-node requests, are not eligible for rewards.

Routability and Ping Latency¶

The visor must be route-able over existing transports and respond to pings. There is no minimum latency requirement, as the measurement is relative to where it is being measured from.

Transport latency is measured automatically when transports are established. If the initial measurement fails (e.g., due to a busy remote visor), the visor will retry with exponential backoff until a measurement succeeds or the transport is closed. Latency data is reported to the transport discovery service on transport re-registration.

Transport Bandwidth and Metrics¶

Transport bandwidth and latency metrics are collected by the transport discovery service. Each visor maintains per-transport bandwidth counters in a local bbolt-backed stats store and reports them when transports are re-registered with the transport discovery. (The on-disk YYYY-MM-DD.csv per-day log store the visor used to keep — and the reward system used to fetch over dmsg — has been retired; historical data is served from the same bbolt store via the visor's RPC.) These metrics can be viewed with:

skywire cli tp metrics
skywire cli tp metrics -t
skywire cli tp metrics --tree

The bandwidth data from the transport discovery is used for the bandwidth pool reward distribution under a sender-pays model: each visor is credited only for the bytes it sent on each transport, and each side's sent counter is capped by the counterparty's reported recv (verified bandwidth). This sender-side cap prevents any single visor from unilaterally inflating its own bandwidth figures.

Survey¶

On setting the skycoin reward address, the visor will generate and serve a system survey over dmsg.

Only keys which are whitelisted in the survey_whitelist array of the visor's config will have access to collect the survey.

To print the survey (approximately) as it would be served, one can use:

skywire cli survey -p

The purpose of the survey is strictly for checking eligibility requirements numbers 3 through 7.

Setting a skycoin address is considered as consent for collecting the survey ; the survey is not generated unless you set a skycoin address

We respect your privacy.

Verifying Other Requirements¶

If the visor is not able to meet the eligibility requirements numbers 8 through 13, that is usually not the fault of the user nor is it something the user is expected to troubleshoot on their own at this time. Please ask for assistance on telegram @skywire

Reward System Overview¶

The skycoin reward address may be set for each visor.

The skycoin reward address is in a text file contained in the "local" folder (local_path in the skywire config file) i.e local/reward.txt.

The skycoin reward address is also included with the system hardware survey and served via dmsghttp.

The system survey ('local/node-info.json') is fetched hourly by the reward system via

skywire cli log

The index of the collected files may be viewed at theskywirenetwork.net/log-collection/tree

Once collected from the nodes, the surveys for those visors which met uptime are checked to verify hardware and other requirements, etc.

The system survey is only made available to those keys which are whitelisted for survey collection, but is additionally available to any hypervisor or dmsgpty_whitelist keys set in the config for a given visor.

Setting a skycoin address is considered as consent for collecting the survey.

The public keys which require to be whitelisted in order to collect the surveys, for the purpose of reward eligibility verification, should populate in the visor's config automatically when the config is generated.

Reward System Funding & Distributions¶

The reward system is funded on a monthly basis. Sometimes there are unexpected or unavoidable delays in the funding. In these instances, rewards will be distributed based on the data generated when the system is funded. In some instances, it's necessary to discard previous reward data and do multiple distributions to handle backlog of reward system funds.

We do our best to ensure fair reward distribution - but the system itself is not infinitely flexible. If there is no good way to rectify historical or undistributed rewards backlog, it will be distributed going forward as multiple distributions on the same day to those current active participants in the network.

Deployment Outages¶

While we do our best to maintain the skywire production deployment, there have been instances of issues or outages in the past. We attempt to correct these outages as soon as possible and avoid recurrent disruptions.

The policy for handling rewards in the instance of a deployment outage is to repeat the distribution for the last day where uptime was unaffected by the outage; for the duration of the outage.

Hardware¶

Skywire rewards are open to all computer hardware and architectures.

If there is not a release for your desired architecture, we can attempt to add it, on request.