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
Total Findings: 19,259 (zizmor 16,484 · grype 2,741 · runner-guard 19 · poutine 7 · actionlint 4 · shellcheck 4 · yamllint 0). actionlint's 4 findings are the same shellcheck notes surfaced through actionlint, so they overlap with the shellcheck row.
Firewall-owned images (agent, api-proxy, cli-proxy, router) carry no Critical/High findings. The Critical load sits in third-party images (serena, xberg, grafana) and in gh-aw-mcpg/squid/gh-aw-node, where bumping the base image (zlib 1.4-r0, golang.org/x/net 0.60.0) clears several at once.
Description: Workflows triggered by comment/issue events with write permissions. These have been reported on every scan since at least 2026-11-07 and are the result of the slash-command/moderation design (role gating and activation checks are applied in the generated activation job). ai-moderator redirects to githubnext/agentics and squad has a source: entry, so both are upstream-managed and not editable here.
Impact: Third-party/base-image CVEs in images pulled by workflows; several share a single remediation (zlib 1.4-r0, x/net 0.60.0, openssl 3.5.8-r0).
3. artipacked (92 findings, 90 workflows)
Tool: zizmor, Medium
Description: The safe-outputs job checks out the repository with persist-credentials: true because it later pushes branches; this is an intentional trade-off in compiler-generated code.
Reference: (docs.zizmor.sh/redacted)
Fix Suggestion for SC2004 (selected and implemented)
Issue: $ on variables used as an indexed-array subscript inside $(( )). Severity: style Affected Workflows: 1 (daily-squid-image-scan)
Prompt to Copilot Agent:
You are fixing a shellcheck SC2004 note surfaced by actionlint.
Rule: SC2004 - $/${} is unnecessary on arithmetic variables
(www.shellcheck.net/redacted)
Problem: inside arithmetic contexts (including array subscripts) bash evaluates bare names, so the dollar sign is redundant.
Fix: edit the workflow .md (never the .lock.yml), change
finding_counts[$match_index]=$((finding_counts[$match_index] + 1))
to
finding_counts[match_index]=$((finding_counts[match_index] + 1))
then run `gh aw compile <workflow> --actionlint` and confirm no issues remain.
Historical Trends
Previous Scan: 2026-11-08
Total zizmor findings then / now: 16,389 / 16,484 (+95, +0.6%; workflows 362 → 365)
Grype total then / now: 2,743 / 2,741 (−2); Critical 79 → 77, High 715 → 713. A lower count alone does not prove a specific CVE was remediated.
Runner-guard: 19 medium RGS-005 (unchanged); no High/Critical since 2026-11-08.
SC2004 in daily-squid-image-scan (fix staged in PR).
Resolved Issues
None confirmed.
Recommendations
Immediate: Review the 3 High dangerous-triggers/bot-conditions workflows; squad and ai-moderator need upstream changes. Bump base images for gh-aw-mcpg, squid, gh-aw-node (zlib, x/net, docker-cli) where fixes exist.
Short-term: Merge the SC2004 PR (maintainer must re-run gh aw compile for locks). Decide whether secrets-outside-env (12.6k) and self-repository (3.4k) should be suppressed in the scanner config, since they dominate noise from compiler-generated steps.
Long-term: Track Critical grype findings per image with owners; consider pinning updated serena/xberg/grafana images.
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.
🔍 Static Analysis Report - 2026-11-09
Analysis Summary
ai-moderator,q,squad).daily-squid-image-scan.md(no Critical/High runner-guard findings, so no runner-guard PRs).Findings by Tool
Clustered Findings by Tool and Type
Zizmor Security Findings
persist-credentials: true)smoke-update-cross-repo-pr(4); informational:smoke-azure-codex,smoke-azure-pi,smoke-service-portsai-moderator,q,squadsquad(line 3045)draft-pr-cleanup,stale-pr-cleanup,super-linterwindows-growersmoke-codexPoutine Supply Chain Findings
copilot-setup-steps(setup-uv),daily-news,org-health-report,audit-workflows,copilot-session-insights(setup-uv),dataflow-pr-discussion-dataset(safedep/pmg) — all pinned to SHAs and carryzizmor: ignorecommentssmoke-copilot-arm(ubuntu-24.04-arm; a GitHub-hosted ARM runner label, likely a false positive)Actionlint Linting Issues
daily-squid-image-scan,sergoSyft SBOM Inventory
Runner-Guard Taint Analysis Findings
Runner-Guard Score: not reported in this output.
ai-moderator(7),q(5),squad(5),agentic-workqueue.yml(1),agentic_commands.yml(1)No Critical/High runner-guard findings → no runner-guard PRs. Pull requests staged: one non-runner-guard fix (see below).
Grype Container Vulnerability Findings
All 12 scans reported explicit completion; none failed or were omitted. Finding totals sum to 2,741.
Per-image scan status
Critical findings by package (77 total; top groups):
Firewall-owned images (agent, api-proxy, cli-proxy, router) carry no Critical/High findings. The Critical load sits in third-party images (serena, xberg, grafana) and in
gh-aw-mcpg/squid/gh-aw-node, where bumping the base image (zlib 1.4-r0, golang.org/x/net 0.60.0) clears several at once.Yamllint YAML Linting Findings
None (0 findings across 365 files).
Shellcheck Shell Linting Findings
daily-squid-image-scan— fixed in staged PRsergo(generated Execute Claude Code CLI step)Top Priority Issues
1. dangerous-triggers + bot-conditions + RGS-005 (slash-command / moderation workflows)
ai-moderator,q,squadactivationjob).ai-moderatorredirects togithubnext/agenticsandsquadhas asource:entry, so both are upstream-managed and not editable here.2. Critical container CVEs with available fixes
3. artipacked (92 findings, 90 workflows)
persist-credentials: truebecause it later pushes branches; this is an intentional trade-off in compiler-generated code.Fix Suggestion for SC2004 (selected and implemented)
Issue:
$on variables used as an indexed-array subscript inside$(( )).Severity: style
Affected Workflows: 1 (
daily-squid-image-scan)Prompt to Copilot Agent:
Historical Trends
github_action_from_unverified_creator_used28 → 6.New Issues
daily-squid-image-scan(fix staged in PR).Resolved Issues
Recommendations
dangerous-triggers/bot-conditionsworkflows;squadandai-moderatorneed upstream changes. Bump base images forgh-aw-mcpg,squid,gh-aw-node(zlib, x/net, docker-cli) where fixes exist.gh aw compilefor locks). Decide whethersecrets-outside-env(12.6k) andself-repository(3.4k) should be suppressed in the scanner config, since they dominate noise from compiler-generated steps.make recompilesurfaces actionlint/shellcheck findings).Next Steps
Run details
squad.md(upstream-managed viasource:) and generated.lock.ymlcontent.All reactions