WIP: Add LWS threat model document - #202
Conversation
|
Awesome @renyuneyun 🎉 I'm going to add few more things to this PR branch tomorrow. @acoburn can we allocate 5 min on Monday to introduce this work during the WG call? We could also use some pointers to how PR previews are setup for this repo.
That lib/template already has an issue to track it moving to W3C github org |
| Data Flow Diagram for LWS Threat Model | ||
| </h3> | ||
| <figure id="lws-threat-model"> | ||
| <img height="1890" src="minimalist-web-threat-model.png" |
There was a problem hiding this comment.
Tomorrow I will replace it with https://hackers4peace.github.io/likec4-example-customization/#lws
|
A preview of this document is at https://htmlpreview.github.io/?https://github.com/renyuneyun/lws-protocol/blob/threat-model/threat-model/index.html We will want to adjust the formatting to bring it in line with W3C conventions |
Thanks @acoburn for having the preview. But... It appears the preview only included the html, without css and js? That leads to incorrect rendering, and incomplete content (of the threats, which are based on js). |
|
This one seems to work better: https://raw.githack.com/renyuneyun/lws-protocol/threat-model/threat-model/index.html |
|
If, on Monday, we merge I'll rebase this PR to reuse the same LikeC4 model and after few updates we could try to vote on this one the following week. |
| <p class="note" role="note"><span class="marker">note:</span>We | ||
| do NOT model potential attackers, as over-characterizing | ||
| attackers can lead to analysis bias that is better avoided.</p> | ||
| did not yet model potential attackers apart from the actors already existed, |
There was a problem hiding this comment.
| did not yet model potential attackers apart from the actors already existed, | |
| did not yet model potential attackers apart from the actors that already existed, |
There was a problem hiding this comment.
Thank you @TallTed, I will double check if we plan to keep that note before applying your suggestion.
The original text comes from this template https://github.com/LegReq/respec-threats/blob/5dd39f65e54221af9dab9a87d635d8cf536c20e9/index.html#L131-L133
| name: "CID and Environment Verification (from User Authorization Server)", | ||
| type: "Reduce", | ||
| desc: "The User Authorization Server verifies the identity of the Client, by checking the environmental information (e.g. redirect URL, Origin) of the Client matches those specified in the Client CID Document.", | ||
| desc: "The User Authorization Server verifies the identity of the Client, by checking the environmental information (e.g. redirect URL) of the Client matches those specified in the Client CID Document. This ensures that the response that the User Authorization Server sends is to a location trusted by the claimed Client CID Document. As long as those specified locations are not compromised, there will not be impersonations.", |
There was a problem hiding this comment.
| desc: "The User Authorization Server verifies the identity of the Client, by checking the environmental information (e.g. redirect URL) of the Client matches those specified in the Client CID Document. This ensures that the response that the User Authorization Server sends is to a location trusted by the claimed Client CID Document. As long as those specified locations are not compromised, there will not be impersonations.", | |
| desc: "The User Authorization Server verifies the identity of the Client, by checking the environmental information (e.g., redirect URL) of the Client matches those specified in the Client CID Document. This ensures that the response sent by the User Authorization Server is to a location trusted by the claimed Client CID Document. As long as those specified locations are not compromised, there will be no impersonations.", |
|
This was discussed during the #lws meeting on 27 July 2026. View the transcriptDiscussion: Add Threat Model document PR#202<ericP> elf-pavlik: since we want to model on top of 159, we're not ready to merge 202 <ericP> ... i will rebase on top of 159 <ericP> PROPOSED: Add Threat Model document PR#202 <gb> Issue 202 not found ericP: anyone thinks they will vote agains? TallTed: given it wasn't on agenda as a vote I'll rather push that ericP: anything more to dicuss on it today? <ericP> elf-pavlik: nope, just everyone take a look (before monday morning) |
Initial threat model covering three threats: - T1: Injected Identity Providers - T2: Impersonation Through Credentials - T3: Client ID Impersonation Based on the respec-threats rendering library (Apache 2.0).
- T2: changed affected elements to P3, P4, P5 - T3: revised threat description and responses, changed elements to P3, P2
Thanks @TallTed
|
Latest version extends C4 from By adding a |
6 participants
Work In Progress
Creating this PR to get discussion going.
[preview]
Info
Initial threat model covering three threats:
More to be added.
Question
Based on the respec-threats rendering library (Apache 2.0). Does that have any copyright issues with LWS Spec when it comes to maturity?