Repository navigation
Expand file tree
/
Copy pathversions.yaml
More file actions
163 lines (146 loc) Β· 5.52 KB
/
Copy pathversions.yaml
File metadata and controls
163 lines (146 loc) Β· 5.52 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
# Single source of truth for every tool version this course pins.
# Lab specs and the README must agree with this file.
#
# Cadence: re-pin everything before each cohort (August, January). During a
# semester a pin changes only when the current one is broken or wrong, and the
# lab's Common pitfalls says so. A student on lab 7 must not find lab 4 changed
# under them.
#
# `verified` records what was actually executed against the pin, and when.
# "recorded" means the pin matches what a student installing today receives, but
# the lab's own commands were not re-run against it.
#
# Check drift: python3 tools/check-versions.py
# Verify a lab: tools/verify-lab.sh 2 --run
checked: 2026-09-06
tools:
juice-shop:
pin: v20.0.0
labs: [1, 4, 5, 7, 8, 9, 10, 11]
github: juice-shop/juice-shop
docker: bkimminich/juice-shop
verified: 2026-09-05, lab 1 commands run end to end
hold: |
v20.2.0 is out and was checked on 2026-09-05: same /api/Products count (46),
same /rest/products 500, same missing headers, same health endpoint. It only
changes the challenge counts (112 -> 116 total, 13 -> 14 Injection) quoted in
Lecture 1. Not bumped because the cohort is on lab 1 with v20.0.0 right now.
Bump it with the lecture counts before the next cohort.
threagile:
pin: "0.9.1"
labs: [2]
github: Threagile/threagile
docker: threagile/threagile
verified: 2026-09-05, lab 2 commands run end to end; latest upstream release
gitleaks:
pin: v8.30.1
labs: [3]
github: gitleaks/gitleaks
verified: 2026-09-05, detects a planted AWS key on a no-git source tree
syft:
pin: v1.51.1
labs: [4]
github: anchore/syft
verified: >
2026-09-05 and 2026-09-06: SBOM of the pinned v20.0.0 image, 3069 CycloneDX
components and 909 SPDX packages (the 20.2.0 image gives 2677)
note: emits CycloneDX specVersion 1.7 by default; Cosign 3.0.2 attests it unchanged
grype:
pin: v0.118.0
labs: [4, 10]
github: anchore/grype
verified: 2026-09-05, scanned the Syft SBOM, 11 critical / 48 high
trivy:
pin: v0.74.0
labs: [4, 6, 7, 10]
github: aquasecurity/trivy
verified: 2026-09-05, image scan of Juice Shop, 50 critical+high
semgrep:
pin: "1.176.0"
labs: [5, 10]
github: semgrep/semgrep
verified: >
recorded. Lab 5 installs it with pip, which gives this release; the docker
`latest` tag trailed it by one patch on 2026-09-05
zap:
pin: v2.17.0
labs: [5, 10]
github: zaproxy/zaproxy
docker: ghcr.io/zaproxy/zaproxy
verified: recorded
hold: |
Labs run the floating `stable` docker tag, so students already get this. The
authenticated scan in lab 5 takes 5-10 minutes and has not been re-run against
2.17; do that before relying on the lab 5 numbers.
checkov:
pin: "3.3.16"
labs: [6, 10]
github: bridgecrewio/checkov
verified: >
2026-09-05, scanned the shipped terraform: still a JSON array of frameworks
(terraform + secrets), severity still null on the OSS build, 78 failed checks
kics:
pin: latest
labs: [6]
docker: checkmarx/kics
verified: recorded
hold: unpinned; pin it to a released tag before the next cohort
k3d:
pin: v5.9.0
labs: [7]
github: k3d-io/k3d
verified: recorded
hold: >
cluster creation and the lab 7 Pod Security Standards steps have not been
re-run against 5.9. Do that before the next cohort.
conftest:
pin: v0.69.0
labs: [7, 9]
github: open-policy-agent/conftest
verified: >
2026-09-05, shipped policies against both lab 9 manifests: hardened 30/30 pass,
unhardened 8 failures, as the lab states
cosign:
pin: v3.0.2
labs: [8]
github: sigstore/cosign
verified: >
2026-09-05, full lab 8 chain: key generation, sign and verify a digest in a
local registry, CycloneDX attestation and verify-attestation round trip
hold: |
Do not move to 3.1.x. It removed support for `--tlog-upload=false` and answers
with a `--signing-config` error, which breaks every signing command in lab 8.
Rewriting the lab for signing configs is the prerequisite for that bump.
falco:
pin: "0.43.1"
labs: [9]
github: falcosecurity/falco
verified: recorded
verified: >
2026-09-06 on a WSL2 kernel with BTF: both built-in detections fired, a
custom rule reloaded with SIGHUP fired, and an outbound-port rule fired on
a refused connection
hold: >
0.44.1 is out and needs the same run before the bump.
defectdojo:
pin: "2.58.3"
labs: [10]
github: DefectDojo/django-DefectDojo
verified: >
2026-09-06, full lab 10 run: stack up in release mode, product and
engagement created through the API, all seven parser names confirmed
present, and the shipped importer loading real Semgrep (26 findings) and
Checkov (80 findings) reports
hold: |
3.x is out. The lab clone is now pinned to 2.58.3 because 3.x renames
importer parsers, which is exactly what run-imports.sh depends on. The bump
needs another full lab 10 run against the new parser names.
kata:
pin: "4.1.0"
labs: [12]
github: kata-containers/kata-containers
verified: >
2026-09-06, release asset kata-static-4.1.0-amd64.tar.zst resolves and the
shipped installer is bash 3.2 clean. The on-host install itself (containerd
runtime registration, /opt/kata) was not executed: it modifies the machine
it runs on and needs /dev/kvm plus containerd rather than Docker alone.