You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[uk ai resilience] UK AI Open Code Risk & Resilience Report – 2026-11-13
#77499
Lookback 7 days: 363 commits (208 with security signals), 24 open code-scanning alerts, 0 secret-scanning alerts, 26 open security-labelled issues. Secret exposure controls look healthy. Main gaps: 8 high-severity CWE-59 (symlink/file-write) AI-scan alerts in actions/setup/sh/*, non-deterministic installs in compiled lock workflows, and aging container CVE burn-down issues (oldest 2026-10-13).
Caveat: sub-agents and some shell reads were unavailable under tool permissions, so scoring relies on the pre-computed alert and issue data plus limited file reads. Confidence is medium.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Executive summary
Lookback 7 days: 363 commits (208 with security signals), 24 open code-scanning alerts, 0 secret-scanning alerts, 26 open security-labelled issues. Secret exposure controls look healthy. Main gaps: 8 high-severity CWE-59 (symlink/file-write) AI-scan alerts in
actions/setup/sh/*, non-deterministic installs in compiled lock workflows, and aging container CVE burn-down issues (oldest 2026-10-13).Caveat: sub-agents and some shell reads were unavailable under tool permissions, so scoring relies on the pre-computed alert and issue data plus limited file reads. Confidence is medium.
Asset graph summary (recent-change scoped)
Asset graph
actions/setup/shactions/setup/jspkg/cli.github/workflows/*.lock.ymlcopilot-maintenance.yml,upgrade-test.ymlscripts/install-gh-aw.shOwnership signals were not verified per path (no CODEOWNERS check).
Tier classification
Tier table
No Tier D candidates identified.
Control verification gaps
start_mcp_gateway.sh:460(1088).Risk scoring
Scores (1 low risk – 5 high risk)
Rationale: the shell helpers run in every compiled workflow, amplifying exposure. Alerts exist but are untriaged.
Remediation queue
actions/setup/sh(1089-1094)Human review triggers: changes to patch generation or usage artifact collection; any ownership gap.
Exception register
None. Any exception needs threat hypothesis, exploit acceleration claim, operational weakness, expiry date and mitigation plan.
Operational metrics baseline
All reactions