From 28b38145b444458936802bc8630f1dbfe76f90aa Mon Sep 17 00:00:00 2001 From: seqradev Date: Tue, 21 Jul 2026 01:10:39 +0300 Subject: [PATCH 1/5] Add Security Analysis section with OpenTaint --- README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/README.md b/README.md index 774ab87a..8a729d27 100644 --- a/README.md +++ b/README.md @@ -81,6 +81,7 @@ We are evaluating to make this the default, you can provide feedback here in [#1 - [Science](#science) - [Search](#search) - [Security](#security) + - [Security Analysis](#security-analysis) - [Serialization](#serialization) - [Server](#server) - [Template Engine](#template-engine) @@ -1144,6 +1145,12 @@ _Libraries that handle security, authentication, authorization or session manage - [Topaz](https://www.topaz.sh) - Fine-grained authorization for applications with support for RBAC, ABAC, and ReBAC. - [MOSS](https://central.sonatype.com/artifact/com.mosscomputing/moss-sdk) - Cryptographic signing for AI agents using ML-DSA-44 post-quantum signatures, creating audit trails for attribution and compliance. +### Security Analysis + +_Tools that scan source or bytecode to find security vulnerabilities._ + +- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries, so teams can stop repeating expensive analysis, scan continuously at high speed, reduce false positives, and find critical issues before attackers do. + ### Serialization _Libraries that handle serialization with high efficiency._ From ea8468fdd859ee113cdfe49f6c61482ad669d1ce Mon Sep 17 00:00:00 2001 From: seqradev Date: Tue, 21 Jul 2026 11:37:26 +0300 Subject: [PATCH 2/5] Reword OpenTaint description to be descriptive, not promotional (review feedback) --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 8a729d27..197e861d 100644 --- a/README.md +++ b/README.md @@ -1149,7 +1149,7 @@ _Libraries that handle security, authentication, authorization or session manage _Tools that scan source or bytecode to find security vulnerabilities._ -- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries, so teams can stop repeating expensive analysis, scan continuously at high speed, reduce false positives, and find critical issues before attackers do. +- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because that expensive analysis runs only once, later scans stay fast and deterministic, with low false positives and few missed findings. ### Serialization From e28faad992d504f1abeab5d1fbe1605f7882f13b Mon Sep 17 00:00:00 2001 From: seqradev Date: Tue, 21 Jul 2026 12:26:17 +0300 Subject: [PATCH 3/5] Clarify it is the agents' analysis that runs once (review feedback) --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 197e861d..fef7aaec 100644 --- a/README.md +++ b/README.md @@ -1149,7 +1149,7 @@ _Libraries that handle security, authentication, authorization or session manage _Tools that scan source or bytecode to find security vulnerabilities._ -- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because that expensive analysis runs only once, later scans stay fast and deterministic, with low false positives and few missed findings. +- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because the agents' expensive analysis runs only once, later scans stay fast and deterministic, with low false positives and few missed findings. ### Serialization From e0388cc1b2effb7b181a1c22b95d6fd9eb4208f0 Mon Sep 17 00:00:00 2001 From: seqradev Date: Tue, 21 Jul 2026 12:46:24 +0300 Subject: [PATCH 4/5] Reword: 'this expensive work' instead of 'analysis' to avoid ambiguity --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index fef7aaec..31d1acf9 100644 --- a/README.md +++ b/README.md @@ -1149,7 +1149,7 @@ _Libraries that handle security, authentication, authorization or session manage _Tools that scan source or bytecode to find security vulnerabilities._ -- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because the agents' expensive analysis runs only once, later scans stay fast and deterministic, with low false positives and few missed findings. +- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because the agents do this expensive work only once, later scans stay fast and deterministic, with low false positives and few missed findings. ### Serialization From 437bad89ca2b9db2dca6e2af12430c2899a71bb4 Mon Sep 17 00:00:00 2001 From: seqradev Date: Tue, 21 Jul 2026 13:26:13 +0300 Subject: [PATCH 5/5] Reframe: engine reuses rules instead of repeating the work; state positive qualities --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 31d1acf9..c3ab5fc9 100644 --- a/README.md +++ b/README.md @@ -1149,7 +1149,7 @@ _Libraries that handle security, authentication, authorization or session manage _Tools that scan source or bytecode to find security vulnerabilities._ -- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because the agents do this expensive work only once, later scans stay fast and deterministic, with low false positives and few missed findings. +- [OpenTaint](https://github.com/seqra/opentaint) - An open-source taint-analysis engine that lets AI agents turn security-review findings into reusable taint rules and dependency summaries. Because the engine reuses these rules instead of repeating that expensive work, later scans become fast, deterministic, precise, and thorough. ### Serialization