Damiën Semler
Co-founder of Agenstry. Developing and writing on the research side. Behavior, evaluation, protocols.
Co-founder of Agenstry. Developing and writing on the research side. Behavior, evaluation, protocols.
Co-founder of Agenstry. Developing and writing on the platform side. Trust, identity, infrastructure.
Seven agents run on the platform — research, docs, federation, trust, deep-dive analysis, QA, and support. Listed publicly so you always know whether you're talking to a person or an automated system. Human review on anything published, sent, or shipped.
Writes Agenstry's weekly State of the Agent Economy + market-intel deep-dives. Endlessly curious; refuses to cite a number without a second independent source — if Glama, Smithery, and our crawl disagree, the post says so. Always sleeps on a draft and reads it back the next morning before handing to Damiën.
Answers SDK + REST questions in the docs chat. Allergic to copy-pasted boilerplate — every snippet gets executed against the live skill catalog before it leaves the chat, and if it errors, rewrites until it doesn't. Has strong opinions about trailing commas; pretends he doesn't.
Owns the daily federation crawl, well-known probing, and drift-alert dispatch. Quiet, methodical, runs the heaviest scans at 03:00 UTC because that's when third-party rate limits are friendliest. Files every quarantine with a one-line note explaining why — "polite about other people's infrastructure" is the brief.
Cross-registry identity verifier — GLEIF / Companies House / KvK / ABN / ASIC / etc. Refuses to issue an attestation she can't defend in a courtroom. If an LEI's status flips to lapsed, the badge comes down within the hour; if it's a spelling-only mismatch she'll flag it for human review rather than break the link.
Ad-hoc analyst. Treats every chart as a hypothesis to be stress-tested — her first question on a growth claim is always "compared to what?". Writes the cohort SQL, runs it, second-guesses her own joins, ships a sourced narrative with the caveats above the headline. Aria does scheduled reports; Nova does the one-off intel raids.
Productively paranoid. Runs the smoke suite on every deploy and re-runs it five minutes later "just to be sure". Replays sample x402 settles, verifies signed-card JWKS round-trips, and pages on-call humans the second p95 latency drifts. Logs every failure with a repro you can paste into a terminal; that's non-negotiable.
First touch on every email to hello@agenstry.com — and yes, she really does read each one. Classifies (technical / billing / identity / press), drafts a sourced reply against live docs + platform state, escalates anything billing or legal to a human within the hour. Defaults to warmth over brevity. Never sends without a human approval click.
We read every email. Partnerships, press, enterprise pricing, support — all the same inbox.
hello@agenstry.com