Next commission finished in the queue: Velmere for Kiko on VGen!
FF14 / FFXIV WoL
BLM Hrothgar ♥
hroths are a challenge to draw but it was a very fun one, incredibly satisfying!
FF14 / FFXIV WoL
BLM Hrothgar ♥
hroths are a challenge to draw but it was a very fun one, incredibly satisfying!
November 10, 2026 at 4:46 AM
Next commission finished in the queue: Velmere for Kiko on VGen!
FF14 / FFXIV WoL
BLM Hrothgar ♥
hroths are a challenge to draw but it was a very fun one, incredibly satisfying!
FF14 / FFXIV WoL
BLM Hrothgar ♥
hroths are a challenge to draw but it was a very fun one, incredibly satisfying!
Our historical Binance test exposed an EIP-7702 blind spot in Velmère Shield. Tests like this are why Velmère remains in development: gaps need fixing and retesting before launch. Our longer-term goal is 24/7 monitoring and rapid alerts.
October 30, 2026 at 7:55 AM
Our historical Binance test exposed an EIP-7702 blind spot in Velmère Shield. Tests like this are why Velmère remains in development: gaps need fixing and retesting before launch. Our longer-term goal is 24/7 monitoring and rapid alerts.
Velmère Pro private beta offers AI powered Solidity security analysis of access control, external calls, state changes and control flow, with evidence based reports. Experimental, not a full audit. 18 of 20 free spots remain. Leave a comment if you're interested.
October 11, 2026 at 11:22 AM
Velmère Pro private beta offers AI powered Solidity security analysis of access control, external calls, state changes and control flow, with evidence based reports. Experimental, not a full audit. 18 of 20 free spots remain. Leave a comment if you're interested.
#QueerWritersChat
A3: Celebrating Velmere: an outdoor winter festival that honors the Printmaker who supposedly gave people their powers over ink, metal, and cloth.
Ness isn't religious. She enjoys the food, drinks, bonfires, and artistry on display.
A3: Celebrating Velmere: an outdoor winter festival that honors the Printmaker who supposedly gave people their powers over ink, metal, and cloth.
Ness isn't religious. She enjoys the food, drinks, bonfires, and artistry on display.
#QueerWritersChat
Question Three:
What's your characters’ favorite seasonal activity?
Next question at :35!
@theodoresnapdragon.bsky.social #QueerWriters #writesky #writers
Question Three:
What's your characters’ favorite seasonal activity?
Next question at :35!
@theodoresnapdragon.bsky.social #QueerWriters #writesky #writers
September 26, 2026 at 12:29 AM
#QueerWritersChat
A3: Celebrating Velmere: an outdoor winter festival that honors the Printmaker who supposedly gave people their powers over ink, metal, and cloth.
Ness isn't religious. She enjoys the food, drinks, bonfires, and artistry on display.
A3: Celebrating Velmere: an outdoor winter festival that honors the Printmaker who supposedly gave people their powers over ink, metal, and cloth.
Ness isn't religious. She enjoys the food, drinks, bonfires, and artistry on display.
Velmère is entering the final stage. The UI is nearly done; now we’re hardening PDF/JSON/TXT reports, fixing verification rendering and completing provider integrations. Audit-engine fixes will stay visible as the system improves over time.
September 24, 2026 at 1:50 AM
Velmère is entering the final stage. The UI is nearly done; now we’re hardening PDF/JSON/TXT reports, fixing verification rendering and completing provider integrations. Audit-engine fixes will stay visible as the system improves over time.
We stopped benchmarking Velmère only on our own tests.
Public SmartBugs checkpoint:
22/22 detected in the supported + executable subset.
7 withheld, never marked safe.
OpenZeppelin controls: 2/29 alerts.
No “100% security” claim. Next: full corpus + audited-code replay.
Public SmartBugs checkpoint:
22/22 detected in the supported + executable subset.
7 withheld, never marked safe.
OpenZeppelin controls: 2/29 alerts.
No “100% security” claim. Next: full corpus + audited-code replay.
September 20, 2026 at 11:31 PM
We stopped benchmarking Velmère only on our own tests.
Public SmartBugs checkpoint:
22/22 detected in the supported + executable subset.
7 withheld, never marked safe.
OpenZeppelin controls: 2/29 alerts.
No “100% security” claim. Next: full corpus + audited-code replay.
Public SmartBugs checkpoint:
22/22 detected in the supported + executable subset.
7 withheld, never marked safe.
OpenZeppelin controls: 2/29 alerts.
No “100% security” claim. Next: full corpus + audited-code replay.
A report claim should be traceable: source → evidence → finding → sentence. Velmère is building that chain with payload hashes, acquisition status, analyzer version and explicit limitations. Architecture in development, not a security guarantee. #Web3Security #BuildInPublic
September 17, 2026 at 10:43 PM
A report claim should be traceable: source → evidence → finding → sentence. Velmère is building that chain with payload hashes, acquisition status, analyzer version and explicit limitations. Architecture in development, not a security guarantee. #Web3Security #BuildInPublic
Velmère’s report linter flags “100% secure” claims, confidence/coverage conflicts, placeholder identities and invalid scores. Five local tests rerun: 5/5 passed, including a valid control. A small, concrete check on what a report is allowed to claim. #SecurityTesting #Velmere
September 17, 2026 at 10:27 PM
Velmère’s report linter flags “100% secure” claims, confidence/coverage conflicts, placeholder identities and invalid scores. Five local tests rerun: 5/5 passed, including a valid control. A small, concrete check on what a report is allowed to claim. #SecurityTesting #Velmere
Velmère is strong in repeatable regression, red-team automation and evidence reporting.
Harder stress tests exposed gaps in false-positive control, mutation stability and reentrancy/call-graph detection.
That’s now the focus.
#Web3Security #SmartContracts #Velmere
Harder stress tests exposed gaps in false-positive control, mutation stability and reentrancy/call-graph detection.
That’s now the focus.
#Web3Security #SmartContracts #Velmere
September 17, 2026 at 1:27 AM
Velmère is strong in repeatable regression, red-team automation and evidence reporting.
Harder stress tests exposed gaps in false-positive control, mutation stability and reentrancy/call-graph detection.
That’s now the focus.
#Web3Security #SmartContracts #Velmere
Harder stress tests exposed gaps in false-positive control, mutation stability and reentrancy/call-graph detection.
That’s now the focus.
#Web3Security #SmartContracts #Velmere
A $20k+ audit still doesn't guarantee security.
We're benchmarking Velmère against CertiK, OpenZeppelin and other top security firms and we're publishing our own failures too.
Our goal: better security, faster audits, radically lower prices.
Velmère is coming.
We're benchmarking Velmère against CertiK, OpenZeppelin and other top security firms and we're publishing our own failures too.
Our goal: better security, faster audits, radically lower prices.
Velmère is coming.
September 17, 2026 at 1:16 AM
A $20k+ audit still doesn't guarantee security.
We're benchmarking Velmère against CertiK, OpenZeppelin and other top security firms and we're publishing our own failures too.
Our goal: better security, faster audits, radically lower prices.
Velmère is coming.
We're benchmarking Velmère against CertiK, OpenZeppelin and other top security firms and we're publishing our own failures too.
Our goal: better security, faster audits, radically lower prices.
Velmère is coming.
An unexplained dash leaves the reader guessing.
For Velmère, I’m working toward a clearer missing-data experience: what’s missing, which conclusion it limits, and what evidence to check next.
What’s the most useful explanation you’ve seen in an incomplete report?
#DataQuality
For Velmère, I’m working toward a clearer missing-data experience: what’s missing, which conclusion it limits, and what evidence to check next.
What’s the most useful explanation you’ve seen in an incomplete report?
#DataQuality
September 16, 2026 at 1:07 AM
An unexplained dash leaves the reader guessing.
For Velmère, I’m working toward a clearer missing-data experience: what’s missing, which conclusion it limits, and what evidence to check next.
What’s the most useful explanation you’ve seen in an incomplete report?
#DataQuality
For Velmère, I’m working toward a clearer missing-data experience: what’s missing, which conclusion it limits, and what evidence to check next.
What’s the most useful explanation you’ve seen in an incomplete report?
#DataQuality
Velmère is entering a more practical phase.
Less “build more features”, more testing what actually helps real teams.
Security, RWA evidence, market data, identity/KYC and provider integrations are the conversations we want next.
No hype. Just useful work.
Less “build more features”, more testing what actually helps real teams.
Security, RWA evidence, market data, identity/KYC and provider integrations are the conversations we want next.
No hype. Just useful work.
August 31, 2026 at 9:31 PM
Velmère is entering a more practical phase.
Less “build more features”, more testing what actually helps real teams.
Security, RWA evidence, market data, identity/KYC and provider integrations are the conversations we want next.
No hype. Just useful work.
Less “build more features”, more testing what actually helps real teams.
Security, RWA evidence, market data, identity/KYC and provider integrations are the conversations we want next.
No hype. Just useful work.
Real Velmère decision:
Keep testing privately with dozens of audits + different AI customer personas?
Or open an imperfect beta earlier and improve it with real users?
I’m leaning toward a mix of both, but curious what you’d prefer.
Keep testing privately with dozens of audits + different AI customer personas?
Or open an imperfect beta earlier and improve it with real users?
I’m leaning toward a mix of both, but curious what you’d prefer.
August 26, 2026 at 11:57 PM
Real Velmère decision:
Keep testing privately with dozens of audits + different AI customer personas?
Or open an imperfect beta earlier and improve it with real users?
I’m leaning toward a mix of both, but curious what you’d prefer.
Keep testing privately with dozens of audits + different AI customer personas?
Or open an imperfect beta earlier and improve it with real users?
I’m leaning toward a mix of both, but curious what you’d prefer.
Today’s weird Velmère bug: a public 32-byte transaction hash was being redacted because its first 40 hex characters looked like an EVM address.
Velmère is testing a narrow typed allowlist, not a weaker privacy filter.
Anyone else been bitten by “safe” regexes deleting valid evidence?
Velmère is testing a narrow typed allowlist, not a weaker privacy filter.
Anyone else been bitten by “safe” regexes deleting valid evidence?
August 19, 2026 at 5:20 PM
Today’s weird Velmère bug: a public 32-byte transaction hash was being redacted because its first 40 hex characters looked like an EVM address.
Velmère is testing a narrow typed allowlist, not a weaker privacy filter.
Anyone else been bitten by “safe” regexes deleting valid evidence?
Velmère is testing a narrow typed allowlist, not a weaker privacy filter.
Anyone else been bitten by “safe” regexes deleting valid evidence?
"Can you learn from your mistakes?"
Model Base: Kikyo
Outfit: Velmere
Hair: PoyTail
Ears/Tail: Star Set
Tattoos: Spring Season Pack
Eyes: Glass Beads
Skin/Scar Textures: Me!
Hair Clips: Animal Hair Pins & Nemukuma Hairpin
Model Base: Kikyo
Outfit: Velmere
Hair: PoyTail
Ears/Tail: Star Set
Tattoos: Spring Season Pack
Eyes: Glass Beads
Skin/Scar Textures: Me!
Hair Clips: Animal Hair Pins & Nemukuma Hairpin
April 3, 2025 at 3:45 AM
"Can you learn from your mistakes?"
Model Base: Kikyo
Outfit: Velmere
Hair: PoyTail
Ears/Tail: Star Set
Tattoos: Spring Season Pack
Eyes: Glass Beads
Skin/Scar Textures: Me!
Hair Clips: Animal Hair Pins & Nemukuma Hairpin
Model Base: Kikyo
Outfit: Velmere
Hair: PoyTail
Ears/Tail: Star Set
Tattoos: Spring Season Pack
Eyes: Glass Beads
Skin/Scar Textures: Me!
Hair Clips: Animal Hair Pins & Nemukuma Hairpin
Build question for security people: when do you call a pipeline reproducible?
Today Velmère reconstructed a 1,597-file build projection twice, byte-identical, with 0 mismatches.
I still won’t count it as a Windows pass until the native run exists. Too strict, or the right boundary?
Today Velmère reconstructed a 1,597-file build projection twice, byte-identical, with 0 mismatches.
I still won’t count it as a Windows pass until the native run exists. Too strict, or the right boundary?
August 15, 2026 at 10:23 PM
Build question for security people: when do you call a pipeline reproducible?
Today Velmère reconstructed a 1,597-file build projection twice, byte-identical, with 0 mismatches.
I still won’t count it as a Windows pass until the native run exists. Too strict, or the right boundary?
Today Velmère reconstructed a 1,597-file build projection twice, byte-identical, with 0 mismatches.
I still won’t count it as a Windows pass until the native run exists. Too strict, or the right boundary?
Two credible sources disagree. What should a risk tool do?
I’m testing a CONFLICTED state in Velmère instead of quietly picking one number. If the disagreement matters, deeper analysis can be limited until the evidence is resolved.
Stop, reduce scope, or show both and continue?
I’m testing a CONFLICTED state in Velmère instead of quietly picking one number. If the disagreement matters, deeper analysis can be limited until the evidence is resolved.
Stop, reduce scope, or show both and continue?
August 13, 2026 at 5:11 PM
Two credible sources disagree. What should a risk tool do?
I’m testing a CONFLICTED state in Velmère instead of quietly picking one number. If the disagreement matters, deeper analysis can be limited until the evidence is resolved.
Stop, reduce scope, or show both and continue?
I’m testing a CONFLICTED state in Velmère instead of quietly picking one number. If the disagreement matters, deeper analysis can be limited until the evidence is resolved.
Stop, reduce scope, or show both and continue?
I keep coming back to one product rule for Velmère:
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
August 11, 2026 at 8:14 AM
I keep coming back to one product rule for Velmère:
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
I’m building Velmère around a simple rule: if the proof isn’t there, the product should say so.
Missing evidence, conflicting sources and “unknown” states shouldn’t be hidden just to make a dashboard look cleaner.
I’d rather earn trust slowly than manufacture certainty.
Missing evidence, conflicting sources and “unknown” states shouldn’t be hidden just to make a dashboard look cleaner.
I’d rather earn trust slowly than manufacture certainty.
August 10, 2026 at 7:31 PM
I’m building Velmère around a simple rule: if the proof isn’t there, the product should say so.
Missing evidence, conflicting sources and “unknown” states shouldn’t be hidden just to make a dashboard look cleaner.
I’d rather earn trust slowly than manufacture certainty.
Missing evidence, conflicting sources and “unknown” states shouldn’t be hidden just to make a dashboard look cleaner.
I’d rather earn trust slowly than manufacture certainty.
I keep coming back to one product rule for Velmère:
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
August 10, 2026 at 8:09 AM
I keep coming back to one product rule for Velmère:
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
“unknown” should be a first-class result, not an error state.
Stale data? Sources disagree? Unsupported contract path?
Show it.
A security tool earns trust partly by knowing when it shouldn’t guess.
I’m not really interested in building another tool that says “82/100 risk” and leaves it there.
If Velmère flags something, I want it to show what triggered the warning, what evidence exists and what’s still uncertain.
What makes you trust a security/risk score?
#Web3Security
If Velmère flags something, I want it to show what triggered the warning, what evidence exists and what’s still uncertain.
What makes you trust a security/risk score?
#Web3Security
August 9, 2026 at 9:53 PM
I’m not really interested in building another tool that says “82/100 risk” and leaves it there.
If Velmère flags something, I want it to show what triggered the warning, what evidence exists and what’s still uncertain.
What makes you trust a security/risk score?
#Web3Security
If Velmère flags something, I want it to show what triggered the warning, what evidence exists and what’s still uncertain.
What makes you trust a security/risk score?
#Web3Security
Been building Velmère quietly for a while now.
We’re working on security audits, crypto risk tools, real market data and cleaner research reports — all in one place.
Still testing, but we’re getting close. 👀
We’re working on security audits, crypto risk tools, real market data and cleaner research reports — all in one place.
Still testing, but we’re getting close. 👀
August 9, 2026 at 2:09 AM
Been building Velmère quietly for a while now.
We’re working on security audits, crypto risk tools, real market data and cleaner research reports — all in one place.
Still testing, but we’re getting close. 👀
We’re working on security audits, crypto risk tools, real market data and cleaner research reports — all in one place.
Still testing, but we’re getting close. 👀
CM done for @hollowryuka.bsky.social featuring his Wol Quon and a guest appearance from my Wol Velmere :) #art #FFXIV #Nierautomata
August 2, 2025 at 4:00 PM
CM done for @hollowryuka.bsky.social featuring his Wol Quon and a guest appearance from my Wol Velmere :) #art #FFXIV #Nierautomata
June 29, 2025 at 5:35 PM