[snarfed]maybe doesn’t matter for you if yous is self-contained. bridgy’s won’t be, though, and i haven’t yet figured out how to prevent people from going under the covers and fabricating arbitrary responses
[snarfed]specifically, if the browser extension just fetches from IG and passes the fetched HTML/JSON on to Bridgy, the attack is to forge that IG HTML/JSON and send it directly to Bridgy
[snarfed]i thought about trust on first use, ie give the client a token when they first send an IG profile to create a Bridgy account, but that’s just as easily forged
[snarfed]the only reasonable defense i’ve come up with is to IndieAuth the user, check for bidirectional rel-me on IG, and then include the IndieAuth token in every request. ugh
aaronpkoh huh, yeah i don't have that problem on mine because the user has to indieauth first anyway in order for the extension to be able to send the micropub request
aaronpkin your case I'd expect something similar, where the worst they could do is make bridgy send a webmention to their own site, but not to other sites
[snarfed]technically bridgy already has a u-url that’s off domain, so the spoofing threat is no greater than spoofing directly on their own server, but in practice i think a bunch of people semi special case bridgy
aaronpki guess even aside from the extension details, you'd probably want to avoid someone being able to set up webmentions for someone else's instagram account to arbitrary websites
[snarfed]right, due to its current level of trust. since if the “attack” is just fabricating arbitrary responses with external u-urls on instagram.com, anyone can do that directly, without bridgy
[Rose], [KevinMarks], shoesNsocks, schmudde, [tw2113_Slack_] and [Raphael_Luckom] joined the channel