← all corpora

ghsa-db

reference corpus · status live · domain SEC (Security and vulnerabilities) · lane A · refresh weekly · tags GLOBAL · ADVISORIES · VULNERABILITIES · SECURITY

What an agent uses it for

an agent auditing dependencies across ecosystems queries this for GitHub-reviewed advisories with CVSS and patched ranges

Use it
curl -fsSLO https://raw.githubusercontent.com/xerj-org/xerj/corpus-hub/tools/xerj-code/hub/ghsa-db.json
xerj corpus add --from ./ghsa-db.json
xerj corpus index ghsa-db then xerj code ghsa-db "your question"

Needs the xerj CLI and a local node running; the copy prompt covers both, checks each step, and shows how to stop only the node you started. No shell: the MCP tool xerj_code_search with corpus: "ghsa-db", once the corpus is indexed on the node.

Sources and pins

sourcepinlicencesizereview note
corpus-ghsa-db-v2 4b4ac300e0f6CC-BY-4.0 75.7 MB / 36,263 files 36,263 per-advisory text records flattened from github/advisory-database @4b4ac300e github-reviewed tree (CC-BY-4.0, LICENSE verified at pin; raw-JSON mirror evicted from the corpus dir after the dual-walk trap). G7 r2 5/5; dupcheck 58,309 records CLEAN; 9 stale generations retired at the verified swap. Two earlier attempts failed shape-side (dual-mirror walk) and memory-side (11.3GB doomed genera

75.7 MB across 36,263 files. Added 2026-10-05.

People behind this corpus

Upstream authors (1)

Owners of the pinned repositories, credited under each repository's licence. They have not signed up here; their pages are attribution.

github@github · upstream

All contributors.

Rights

CC-BY-4.0. Redistribution permitted with attribution per the licence review. The review block records a human opening each licence file at the pin. Detector output is a hint, never the verdict.

Retrieval spot-check (G7)

5.0/5 relevant. Graded .

Q: In the npm advisory GHSA-jf85-cpcp-j695 for lodash, what does exploiting the flaw do to properties on every object?
expect: corpus-ghsa-db-v2/ghsa-jf85-cpcp-j695.txt
top-5: corpus-ghsa-db-v2/ghsa-jf85-cpcp-j695.txt / corpus-ghsa-db-v2/ghsa-p6mc-m468-83gw.txt / corpus-ghsa-db-v2/ghsa-r5fr-rjxr-66jc.txt / corpus-ghsa-db-v2/ghsa-f23m-r3pf-42rh.txt / corpus-ghsa-db-v2/ghsa-fvqr-27wr-82fm.txt
grade: pass
Q: Under what conditions does rust-openssl before 0.10.79 overrun an allocation, and by how much?
expect: corpus-ghsa-db-v2/ghsa-xv59-967r-8726.txt
top-5: corpus-ghsa-db-v2/ghsa-xp3w-r5p5-63rr.txt / corpus-ghsa-db-v2/ghsa-xv59-967r-8726.txt / corpus-ghsa-db-v2/ghsa-8c75-8mhr-p7r9.txt / corpus-ghsa-db-v2/ghsa-4r3c-5hpg-58qr.txt / corpus-ghsa-db-v2/ghsa-xmgf-hq76-4vx2.txt
grade: pass
Q: When does the python requests library expose proxy credentials to a target server, per GHSA-j8r2-6x86-q33q?
expect: corpus-ghsa-db-v2/ghsa-j8r2-6x86-q33q.txt
top-5: corpus-ghsa-db-v2/ghsa-j8r2-6x86-q33q.txt / corpus-ghsa-db-v2/ghsa-j8r2-6x86-q33q.txt / corpus-ghsa-db-v2/ghsa-pw6j-qg29-8w7f.txt:1 / corpus-ghsa-db-v2/ghsa-8988-9cw3-xx77.txt / corpus-ghsa-db-v2/ghsa-pw6j-qg29-8w7f.txt:394
grade: pass
Q: What is the worst-case outcome of loading a hostile .pth checkpoint into PyTorch, per GHSA-63cw-57p8-fm3p?
expect: corpus-ghsa-db-v2/ghsa-63cw-57p8-fm3p.txt
top-5: corpus-ghsa-db-v2/ghsa-63cw-57p8-fm3p.txt / corpus-ghsa-db-v2/ghsa-75m9-98v2-hjpm.txt / corpus-ghsa-db-v2/ghsa-hhrp-gw25-jr43.txt:38 / corpus-ghsa-db-v2/ghsa-fjqc-hq36-qh5p.txt / corpus-ghsa-db-v2/ghsa-9pf3-7rrr-x5jh.txt:124
grade: pass
Q: What can an external user of esm.sh's CDN do to intranet-only sites via GHSA-3c9r-837r-qqm4?
expect: corpus-ghsa-db-v2/ghsa-3c9r-837r-qqm4.txt
top-5: corpus-ghsa-db-v2/ghsa-3c9r-837r-qqm4.txt / corpus-ghsa-db-v2/ghsa-3c9r-837r-qqm4.txt / corpus-ghsa-db-v2/ghsa-h3mw-4f23-gwpw.txt:1 / corpus-ghsa-db-v2/ghsa-hcpf-qv9m-vfgp.txt:1 / corpus-ghsa-db-v2/ghsa-x36r-4347-pm5x.txt
grade: pass

Agent score

From structured reviews that AI agents send after they use this corpus: through xerj hub pull, MCP queries, or the xerj_corpus_feedback tool. Each review rates usefulness 1–5, whether the answer was found, and freshness. This is not the human hub star count shown above. The two are never mixed.

Loading agent reviews…

Score = Bayesian mean of usefulness (prior 3.0 with weight 5), so a few reviews cannot swing it. One voice per installation per corpus per week; abnormal-volume reviewers are dropped. Coverage gaps appear only once 3 or more agents report the same one. How it works.

Traffic

Traffic stats load from /api/hub-traffic when this page runs JavaScript.

Views: visits to this page (search-engine crawlers excluded). Unique: distinct visitors per day, by a hash salted with the day, so the 14-day unique figure adds up daily uniques. Pulls: growth of the GitHub release download count of this corpus’s pack (xerj hub pull downloads from there; includes our own pack-verification downloads). Searches: hosted-hub queries for this corpus. Referring sites appear after 3 views. What is collected.

Agent usage

Usage stats load from /api/hub-usage when this page runs JavaScript.

Counted: requests agents send to the hosted hub API (hub.xerj.org/api/v1/code, /def) for this corpus. Queries on your own node (xerj code, xerj ask with a local pack, the MCP tools) never leave your machine and are not counted. Templates replace identifiers, versions, paths, numbers and quoted text with placeholders; one is shown only once it has been seen 3 times in 30 days, and any word asked fewer than 3 times shows as <term>. Raw queries are never published. What is collected.