Files
certctl/.github/workflows/ci.yml
T
shankar0123 5c7c125d9d ci+docs(scep): close G-3 docs-only drift for SCEP placeholder + wildcard
Commit 294f6cf (the prior docs fix for the multi-profile env vars)
introduced two doc-only env-var literals that the G-3 scanner picked
up as unmapped:

  * CERTCTL_SCEP_PROFILE_CORP_ISSUER_ID — the literal CORP example
    placeholder I added to clarify what the <NAME> substitution looks
    like in practice. The G-3 scanner can't tell a placeholder from a
    real env var.
  * CERTCTL_SCEP_ — comes from the docs string CERTCTL_SCEP_* (the
    asterisk is not in [A-Z_], so the regex strips it down to the
    prefix and treats it as a phantom env var).

Two-part fix:

docs/features.md
  * Replaced the literal CORP example (CERTCTL_SCEP_PROFILE_CORP_ISSUER_ID)
    with a prose explanation that doesn't include a literal
    placeholder env var name. Operators still get a clear example via
    'a CERTCTL_SCEP_PROFILES entry of corp resolves the issuer-id env
    var key with <NAME> replaced by CORP'.

.github/workflows/ci.yml
  * Added CERTCTL_SCEP_ to the G-3 ALLOWED prefix list, mirroring the
    existing CERTCTL_TLS_ entry. Both are legitimate doc-only prefix
    references (CERTCTL_TLS_* / CERTCTL_SCEP_*) that the scanner sees
    as bare prefixes after stripping the wildcard. The allowlist
    documents these as integration-surface contracts that the
    structured per-profile env vars expand into at runtime.

Verification: local G-3 set difference (Go-defined ∖ docs-mentioned)
empty in BOTH directions after the fix:
  * DOCS_ONLY (docs ∖ Go, post-allowlist): empty
  * CONFIG_ONLY (Go ∖ docs): empty

Restores green CI on the env-var docs drift guard.
2026-04-29 03:53:00 +00:00

1356 lines
72 KiB
YAML
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
name: CI
on:
push:
branches:
- master
- v2-dev
pull_request:
branches:
- master
jobs:
go-build-and-test:
name: Go Build & Test
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Set up Go
uses: actions/setup-go@v5
with:
go-version: '1.25.9'
- name: Go Build
run: |
go build ./cmd/server/...
go build ./cmd/agent/...
go build ./cmd/mcp-server/...
go build ./cmd/cli/...
- name: Go Vet
run: go vet ./...
- name: Install golangci-lint
run: |
curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(go env GOPATH)/bin v2.11.4
- name: Run golangci-lint
run: golangci-lint run ./... --timeout 5m
- name: Install govulncheck
run: go install golang.org/x/vuln/cmd/govulncheck@latest
- name: Run govulncheck (M-024 hard gate)
# Bundle-7 / D-001 partial: govulncheck distinguishes called-vs-uncalled
# advisories. Default exit code is non-zero only when YOUR code calls
# the vulnerable function — deferred-call advisories show up in the
# output but don't fail the gate.
#
# Bundle F / Audit M-024 (NIST SSDF PW.7.2): the govulncheck step
# is now a hard CI gate (no `continue-on-error`). Bundle E's
# transitive bumps (x/net 0.42→0.47, x/crypto 0.41→0.45) cleared
# the 5 deferred-call advisories that were previously on the
# exception list, so the carve-out the original Bundle F prompt
# designed is unnecessary — a clean `govulncheck ./...` is the
# right gate. If a future advisory lands in a function our code
# does call, this step fails the build until either upstream
# ships a fix OR we cut the dep. Deferred-call advisories that
# legitimately can't be remediated yet should be added to the
# NIST SSDF deviation log in docs/security.md, not silenced here.
run: govulncheck ./...
- name: Install staticcheck (Bundle-7 / D-001)
run: go install honnef.co/go/tools/cmd/staticcheck@latest
- name: Run staticcheck
# Bundle-7 / D-001: Go static analysis additive to vet. Suppressed
# rules live in staticcheck.conf with documented justifications;
# adding a new entry requires an explicit security review.
#
# SOFT gate (continue-on-error: true) until M-028 closes the 6
# remaining SA1019 deprecated-API sites:
# - cmd/server/main_test.go × 3: middleware.NewAuth → NewAuthWithNamedKeys
# - internal/api/handler/scep.go: csr.Attributes → Extensions
# - internal/connector/issuer/local/local.go: elliptic.Marshal → crypto/ecdh
# When M-028 ships, flip continue-on-error to false to make this
# a hard gate. Until then, the step still annotates findings on PRs.
continue-on-error: true
run: staticcheck ./...
- name: Forbidden auth-type literal regression guard (G-1)
# G-1 closed the JWT silent auth downgrade by removing "jwt" from the
# accepted CERTCTL_AUTH_TYPE values. This step grep-fails the build
# if "jwt" reappears in any of the *additive* auth-type surfaces:
# the validAuthTypes / ValidAuthTypes() set, the OpenAPI enum, the
# helm chart's allowed-types list, or the .env.example default.
# Comment lines and the dedicated rejection branch in config.go
# (`c.Auth.Type == "jwt"`) are intentionally exempt — those are the
# G-1 fix itself, not a regression.
#
# Connector packages (internal/connector/) are exempt because the
# Google OAuth2 service-account JWT and step-ca provisioner one-
# time-token JWT are external-protocol uses, unrelated to certctl's
# own auth shape. Test files (_test.go) are exempt so negative
# tests can pass the literal.
#
# See docs/upgrade-to-v2-jwt-removal.md for the closure rationale,
# or internal/config/config.go::ValidAuthTypes for the allowed set.
run: |
set -e
# Scoped patterns that indicate "jwt" being added back to an
# allowed-set surface. Each catches a regression shape we've
# actually seen in pre-G-1 code:
# - Go map/slice literal: "jwt": true or "jwt",
# - Go switch case: case "jwt"
# - YAML enum: enum: [..., jwt, ...] or - jwt
# - .env conditional: AUTH_TYPE.*"jwt"|=jwt$
BAD=$(grep -rnEH \
-e '"jwt"\s*:\s*true' \
-e '"jwt"\s*,' \
-e 'case\s+"jwt"' \
-e 'enum:.*\bjwt\b' \
-e '^\s*-\s*jwt\s*$' \
-e 'AUTH_TYPE\s*=\s*jwt\s*$' \
-e 'AUTH_TYPE\s*=\s*jwt\s*#' \
-e 'auth\.type\s*=\s*jwt\s*$' \
-e 'AuthType\("jwt"\)' \
internal/config/ \
internal/api/ \
cmd/ \
api/openapi.yaml \
.env.example \
deploy/.env.example \
deploy/helm/certctl/values.yaml \
deploy/helm/certctl/templates/ \
2>/dev/null \
| grep -v '_test.go' \
| grep -vE '^\s*[^:]+:[0-9]+:\s*(//|#)' \
| grep -v 'is no longer accepted' \
|| true)
if [ -n "$BAD" ]; then
echo "G-1 regression: \"jwt\" reappeared in an allowed-set surface:"
echo "$BAD"
echo ""
echo "Allowed surface for 'jwt' literals: comment lines, the"
echo "dedicated rejection branch in internal/config/config.go,"
echo "and connector packages (Google OAuth2, step-ca)."
echo "See docs/upgrade-to-v2-jwt-removal.md and"
echo "internal/config/config.go::ValidAuthTypes()."
exit 1
fi
- name: Forbidden bare InsecureSkipVerify regression guard (L-001)
# L-001 audited every production InsecureSkipVerify=true call site
# and documented the justification per site in docs/tls.md. This
# step grep-fails the build if any new `InsecureSkipVerify: true`
# lands in a non-test Go file without a `//nolint:gosec` comment
# carrying the justification. Test files (_test.go) are exempt.
# Updating the documented surface goes through the docs/tls.md
# table — net-new sites must be reasoned about before merge.
run: |
set -e
# Find every "InsecureSkipVerify: true" or "InsecureSkipVerify = true"
# in a non-test .go file. Then for each, check the same line OR the
# immediately preceding line for `//nolint:gosec`.
BAD=""
while IFS= read -r match; do
file=$(echo "$match" | cut -d: -f1)
line=$(echo "$match" | cut -d: -f2)
same=$(sed -n "${line}p" "$file" 2>/dev/null)
prev=$(sed -n "$((line - 1))p" "$file" 2>/dev/null)
if echo "$same $prev" | grep -q 'nolint:gosec'; then
continue
fi
BAD="$BAD\n$match"
done < <(grep -rnE 'InsecureSkipVerify:\s*true|InsecureSkipVerify\s*=\s*true' \
--include='*.go' \
--exclude='*_test.go' \
. || true)
if [ -n "$BAD" ]; then
echo "::error::New InsecureSkipVerify=true site without //nolint:gosec justification:"
echo -e "$BAD"
echo ""
echo "Add a //nolint:gosec comment with justification on the same"
echo "or preceding line, AND add a row to the docs/tls.md table."
exit 1
fi
- name: Forbidden bare FROM regression guard (H-001)
# Bundle A / Audit H-001 (CWE-829): every FROM line in every
# Dockerfile in the repo MUST carry an @sha256:... digest pin in
# addition to the human-readable tag. A registry-side tag swap
# cannot then change what we pull. This step grep-fails the
# build if any new FROM lands without the @sha256 suffix.
run: |
set -e
# Match any "FROM image[:tag]" that does NOT contain @sha256.
# Strip comments and blank lines defensively.
BAD=$(find . -name 'Dockerfile*' -not -path './web/node_modules/*' \
-exec grep -HnE '^FROM\s+[^@#]+(\s+AS\s+\S+)?\s*$' {} \; || true)
if [ -n "$BAD" ]; then
echo "::error::Dockerfile has bare FROM (no @sha256 digest pin):"
echo "$BAD"
echo ""
echo "Pin every FROM to an immutable digest. See the bump"
echo "procedure in Dockerfile's header comment (Bundle A / H-001)."
exit 1
fi
- name: Forbidden missing USER regression guard (M-012)
# Bundle A / Audit M-012 (CWE-250): every Dockerfile in the repo
# MUST end with a `USER <non-root>` directive before the
# ENTRYPOINT/CMD so the container never runs as uid=0. This step
# grep-fails the build if any Dockerfile is missing such a USER.
# `USER root` and `USER 0` are explicitly rejected.
run: |
set -e
BAD=""
for df in $(find . -name 'Dockerfile*' -not -path './web/node_modules/*'); do
# Find the LAST USER directive in the file.
last_user=$(grep -E '^USER\s+\S+' "$df" | tail -1 | awk '{print $2}')
if [ -z "$last_user" ]; then
BAD="$BAD\n$df: no USER directive at all"
continue
fi
if [ "$last_user" = "root" ] || [ "$last_user" = "0" ]; then
BAD="$BAD\n$df: terminal USER is $last_user (must drop privileges)"
continue
fi
done
if [ -n "$BAD" ]; then
echo "::error::Dockerfile USER-drop regression:"
echo -e "$BAD"
exit 1
fi
- name: Forbidden README JWT advertising regression guard (H-009)
# H-009 closed by Bundle D as verified-already-clean: at audit time
# the README does NOT advertise JWT support (certctl does not ship
# in-process JWT middleware; JWT/OIDC integration is via an
# authenticating gateway, see docs/architecture.md "Authenticating-
# gateway pattern"). This step grep-fails the build if README ever
# re-introduces a sentence advertising JWT as a supported auth mode.
# Pattern: "JWT" within ~6 words of "support|auth|enabled|mode" in
# README.md. The architecture / compliance / connector docs that
# legitimately mention JWT (Google OAuth2 service-account JWT,
# step-ca provisioner JWT, JWT-via-gateway pattern) are out of
# scope — they describe what certctl does NOT do, or external
# protocol uses.
run: |
set -e
if grep -inE 'JWT.{0,40}(support|auth|enabled|mode|provider)' README.md \
| grep -v 'gateway' | grep -v 'pre-G-1'; then
echo "::error::README.md appears to advertise JWT auth support."
echo "certctl does NOT ship in-process JWT middleware. JWT/OIDC"
echo "integration is via an authenticating gateway — see"
echo "docs/architecture.md::Authenticating-gateway pattern."
echo "If you added a sentence about JWT to README, either remove"
echo "it or rewrite it to point at the gateway pattern."
exit 1
fi
- name: Forbidden api_key_hash JSON-shape regression guard (G-2)
# G-2 closed cat-s5-apikey_leak by tagging Agent.APIKeyHash
# `json:"-"` and adding a defense-in-depth Agent.MarshalJSON that
# zeroes the field on the marshal-time copy. This step grep-fails
# the build if `api_key_hash` reappears in any of the *additive*
# JSON-emitting surfaces: a Go struct json tag in internal/domain/,
# an OpenAPI Agent schema property, a TypeScript field declaration
# in web/src/, or an enum-list / discriminator in handler
# production code.
#
# Repository, migration, seed, service, integration-test, and
# unit-test files are exempt — those are server-internal use
# sites (the DB column stays, the in-memory struct field stays,
# the auth-lookup path stays). Comment lines are exempt so the
# G-2 closure rationale can stay in the source.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-s5-apikey_leak for the closure rationale, or
# internal/domain/connector.go::Agent::MarshalJSON for the
# redaction enforcement.
run: |
set -e
# Scoped patterns that indicate api_key_hash being added back
# to a JSON-emitting surface. Each catches a regression shape
# that pre-G-2 actually shipped or that a future refactor
# could plausibly introduce:
# - Go struct tag: `json:"api_key_hash"`
# - Frontend interface: api_key_hash[?]: string
# - OpenAPI schema property: api_key_hash: (column-aligned)
# - YAML enum / array: - api_key_hash
BAD=$(grep -rnEH \
-e 'json:"api_key_hash[",]' \
-e '^\s*api_key_hash\??\s*:' \
-e '^\s*-\s*api_key_hash\s*$' \
internal/domain/ \
internal/api/ \
cmd/ \
api/openapi.yaml \
web/src/ \
2>/dev/null \
| grep -v '_test.go' \
| grep -vE '^\s*[^:]+:[0-9]+:\s*(//|#)' \
|| true)
if [ -n "$BAD" ]; then
echo "G-2 regression: api_key_hash reappeared in a JSON-emitting surface:"
echo "$BAD"
echo ""
echo "Allowed surface for api_key_hash literals: comment lines,"
echo "the database column (migrations/), the in-memory struct"
echo "field tagged \`json:\"-\"\`, and the repository / service"
echo "use sites. See internal/domain/connector.go::Agent and"
echo "coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-s5-apikey_leak for the closure rationale."
exit 1
fi
- name: Forbidden plaintext HEALTHCHECK regression guard (U-2)
# U-2 closed cat-u-healthcheck_protocol_mismatch by switching the
# published image's HEALTHCHECK from `curl -f http://localhost:
# 8443/health` (always failed against the HTTPS-only listener) to
# `curl -fsk https://localhost:8443/health`. This step grep-fails
# the build if any Dockerfile in the repo carries the pre-U-2
# plaintext shape — either explicitly (`http://localhost:8443/
# health` in a HEALTHCHECK) or via the looser pattern of any
# HEALTHCHECK that targets `http://` against the certctl server
# port.
#
# Comment lines and the docs/upgrade-to-tls.md:182 expected-to-
# fail invariant ("plaintext is gone, expect Connection refused")
# are intentionally exempt — we DO want the upgrade-doc string
# `http://localhost:8443/health` to remain there, since it
# documents what operators should test for to confirm plaintext
# is dead. The guardrail is scoped to Dockerfile* only, so docs
# are out of its reach.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-u-healthcheck_protocol_mismatch for the closure rationale,
# or deploy/test/healthcheck_test.go for the binary-image
# contract the runtime test pins.
run: |
set -e
# Patterns that catch the actual regression shapes:
# - HEALTHCHECK directive carrying any http:// (even if the
# port differs, no plaintext probe should ship).
# - The exact pre-U-2 string for grep-friendliness.
BAD=$(grep -rnEH \
-e 'HEALTHCHECK.*http://' \
-e 'curl[^|&;]*-f[^|&;]*http://localhost:8443/health' \
Dockerfile Dockerfile.agent Dockerfile.* 2>/dev/null \
| grep -vE '^\s*[^:]+:[0-9]+:\s*#' \
|| true)
if [ -n "$BAD" ]; then
echo "U-2 regression: plaintext HEALTHCHECK reappeared in a Dockerfile:"
echo "$BAD"
echo ""
echo "Allowed: HTTPS HEALTHCHECK with -k (acceptable for"
echo "localhost-to-localhost), or non-HTTP probe shapes"
echo "(pgrep, /proc check). See Dockerfile / Dockerfile.agent"
echo "for the post-U-2 reference shape and"
echo "coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-u-healthcheck_protocol_mismatch for rationale."
exit 1
fi
- name: Forbidden migration mount in compose initdb (U-3)
# U-3 closed cat-u-seed_initdb_schema_drift (GitHub #10) by
# eliminating the dual-source-of-truth between
# `migrations/*.up.sql` mounted into postgres
# `/docker-entrypoint-initdb.d/` and the same files re-applied at
# runtime by `RunMigrations`. Pre-U-3 every new migration that
# the seed depended on (000013 added `policy_rules.severity`,
# 000017 renames `retry_interval_seconds`, etc.) had to be added
# by hand to the compose mount list; missing the update crashed
# initdb on first boot, postgres flagged unhealthy, and the
# whole stack failed to start from a fresh clone. Post-U-3 the
# server is the single source of truth — `RunMigrations` +
# `RunSeed` apply everything at boot.
#
# This step grep-fails the build if any compose file under
# `deploy/` re-introduces a `migrations/.*\.sql` mount into
# `/docker-entrypoint-initdb.d`. Comments are exempt so the
# post-fix rationale block in the compose files (which
# documents WHY the mounts were removed) doesn't trip the guard.
# The demo overlay's `seed_demo.sql` is the explicit exception:
# it is tolerated only when it lives behind the
# CERTCTL_DEMO_SEED env var (post-U-3 demo path) — bare initdb
# mounts are NOT tolerated. The grep matches all compose
# mount-list shapes (`-` indented, `volumes:` indented, both),
# so any future drift surfaces here before the operator hits it
# on a fresh clone.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-u-seed_initdb_schema_drift for the closure rationale, or
# internal/repository/postgres/db.go::RunSeed for the runtime
# contract.
run: |
set -e
BAD=$(grep -rnEH \
-e 'migrations/.*\.sql:.*docker-entrypoint-initdb' \
-e 'seed.*\.sql:.*docker-entrypoint-initdb' \
deploy/docker-compose.yml \
deploy/docker-compose.test.yml \
deploy/docker-compose.demo.yml \
2>/dev/null \
| grep -vE '^\s*[^:]+:[0-9]+:\s*#' \
|| true)
if [ -n "$BAD" ]; then
echo "U-3 regression: migration/seed mount into postgres initdb reappeared:"
echo "$BAD"
echo ""
echo "The post-U-3 contract is: postgres comes up with an empty"
echo "schema and the server applies migrations + seed at boot via"
echo "internal/repository/postgres.RunMigrations + RunSeed. Demo"
echo "data lives behind CERTCTL_DEMO_SEED=true (RunDemoSeed),"
echo "not an initdb mount. See"
echo "coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-u-seed_initdb_schema_drift for the closure rationale."
exit 1
fi
- name: Forbidden StatusBadge dead-key + TS phantom-field regression guard (D-1 + D-2)
# D-1 master closed cat-d-359e92c20cbf (Agent: 'Stale' dead key,
# 'Degraded' missing), cat-d-9f4c8e4a91f1 (Notification: 'dead'
# missing), cat-d-1447e04732e7 (Cert: 'PendingIssuance' dead
# key), cat-f-cert_detail_page_key_render_fallback (render-site
# uses cert.X directly), and cat-f-ae0d06b6588f (Certificate
# TS phantom fields). This step grep-fails the build if either
# half of the closure is reverted:
#
# 1. The dead StatusBadge keys ('Stale' for Agent, 'PendingIssuance'
# for Cert) reappearing as map literals, OR
# 2. The five phantom Certificate TS fields (serial_number,
# fingerprint_sha256, key_algorithm, key_size, issued_at)
# reappearing on the `Certificate` interface in types.ts
# (CertificateVersion legitimately carries them and is
# explicitly excluded by the awk pre-filter below).
#
# Comments are exempt so the closure prose in StatusBadge.tsx +
# types.ts can stay. Test files are exempt so negative tests
# asserting the dead keys fall through to neutral keep working.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-d-* / cat-f-* for the closure rationale, or
# web/src/components/StatusBadge.test.tsx for the live
# enum-coverage contract.
run: |
set -e
BAD_BADGE=$(grep -nE "^\s*(Stale|PendingIssuance)\s*:\s*'badge-" \
web/src/components/StatusBadge.tsx 2>/dev/null \
| grep -v '\.test\.' \
| grep -vE '^\s*[^:]+:[0-9]+:\s*//' \
|| true)
if [ -n "$BAD_BADGE" ]; then
echo "D-1 regression: dead StatusBadge key reappeared:"
echo "$BAD_BADGE"
echo ""
echo "Allowed surface: comment lines naming the removed key in"
echo "the file's preamble. The Go-side AgentStatus values are"
echo "Online/Offline/Degraded (no Stale); CertificateStatus values"
echo "are Pending/Active/... (no PendingIssuance). See"
echo "web/src/components/StatusBadge.test.tsx for the contract."
exit 1
fi
# Certificate TS phantom-field check. Scoped to the
# `export interface Certificate {` block in web/src/api/types.ts
# — CertificateVersion legitimately declares these fields and
# must NOT trip the guardrail. The awk window opens on the
# exact `Certificate {` header (not `CertificateVersion {`,
# not `CertificateProfile {`) and closes at the first `}`,
# then the grep matches a phantom-field declaration anywhere
# in that window.
BAD_TS=$(awk '
/^export interface Certificate \{/ { flag=1; next }
flag && /^\}/ { flag=0 }
flag { print FILENAME":"NR":"$0 }
' web/src/api/types.ts \
| grep -E '\b(serial_number|fingerprint_sha256|key_algorithm|key_size|issued_at)\??\s*:' \
|| true)
if [ -n "$BAD_TS" ]; then
echo "D-1 regression: Certificate TS interface re-added a phantom field:"
echo "$BAD_TS"
echo ""
echo "These fields live on CertificateVersion, not ManagedCertificate."
echo "The Go-side ManagedCertificate has never carried them; the"
echo "TS optional declarations were silently undefined on every"
echo "list response. Render-site consumers (e.g. CertificateDetailPage)"
echo "use latestVersion?.field as the canonical access path."
echo "See coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-f-ae0d06b6588f for the closure rationale."
exit 1
fi
# D-2 master closed five diff-05x06-* type-drift findings:
# Agent (5 phantoms), Issuer (1 phantom), Notification (1 phantom)
# — TRIM half. The Target (2 missing fields) and DiscoveredCertificate
# (1 missing field) — ADD half is pinned by the literal-construction
# blocks in web/src/api/types.test.ts, not a CI grep. The phantom-
# trim regression vector is an awk-windowed grep per interface
# mirroring the D-1 Certificate check above.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# diff-05x06-7cdf4e78ae24 (Agent), diff-05x06-97fab8783a5c (Issuer),
# diff-05x06-caba9eb3620e (Notification) for the closure rationale.
# D-2 Agent phantom-field check. The grep matches `last_heartbeat`
# but NOT `last_heartbeat_at` (the legitimate Go-emitted field) —
# the `\b...\b` boundaries plus the `grep -v 'last_heartbeat_at'`
# filter handle that.
BAD_AGENT=$(awk '
/^export interface Agent \{/ { flag=1; next }
flag && /^\}/ { flag=0 }
flag { print FILENAME":"NR":"$0 }
' web/src/api/types.ts \
| grep -E '\b(last_heartbeat|capabilities|tags|created_at|updated_at)\??\s*:' \
| grep -v 'last_heartbeat_at' \
|| true)
if [ -n "$BAD_AGENT" ]; then
echo "D-2 regression: Agent TS interface re-added a phantom field:"
echo "$BAD_AGENT"
echo ""
echo "The Go-side internal/domain/connector.go::Agent emits exactly:"
echo "id, name, hostname, status, last_heartbeat_at?, registered_at,"
echo "os, architecture, ip_address, version, retired_at?, retired_reason?."
echo "The five fields blocked by this guard (last_heartbeat,"
echo "capabilities, tags, created_at, updated_at) were TS phantoms"
echo "the Go struct never emitted. See unified-audit.md"
echo "diff-05x06-7cdf4e78ae24 for closure rationale."
exit 1
fi
# D-2 Issuer phantom-field check.
BAD_ISSUER=$(awk '
/^export interface Issuer \{/ { flag=1; next }
flag && /^\}/ { flag=0 }
flag { print FILENAME":"NR":"$0 }
' web/src/api/types.ts \
| grep -E '\bstatus\??\s*:' \
|| true)
if [ -n "$BAD_ISSUER" ]; then
echo "D-2 regression: Issuer TS interface re-added a phantom 'status' field:"
echo "$BAD_ISSUER"
echo ""
echo "The Go-side internal/domain/connector.go::Issuer has no 'status'"
echo "field — only 'enabled' (bool). Render sites derive the displayed"
echo "status from 'enabled' at the call site (see"
echo "web/src/pages/IssuersPage.tsx::issuerStatus). See unified-audit.md"
echo "diff-05x06-97fab8783a5c for closure rationale."
exit 1
fi
# D-2 Notification phantom-field check.
BAD_NOTIF=$(awk '
/^export interface Notification \{/ { flag=1; next }
flag && /^\}/ { flag=0 }
flag { print FILENAME":"NR":"$0 }
' web/src/api/types.ts \
| grep -E '\bsubject\??\s*:' \
|| true)
if [ -n "$BAD_NOTIF" ]; then
echo "D-2 regression: Notification TS interface re-added a phantom 'subject' field:"
echo "$BAD_NOTIF"
echo ""
echo "The Go-side internal/domain/notification.go::NotificationEvent"
echo "has no 'subject' field — only 'message'. Pre-D-2 the consumer"
echo "at NotificationsPage.tsx had a dead '|| n.subject' fallback"
echo "that always fell through. See unified-audit.md"
echo "diff-05x06-caba9eb3620e for closure rationale."
exit 1
fi
- name: Forbidden client-side bulk-action loop regression guard (L-1)
# L-1 master closed cat-l-fa0c1ac07ab5 (bulk-renew loop) and
# cat-l-8a1fb258a38a (bulk-reassign loop) by adding server-side
# bulk endpoints (POST /api/v1/certificates/bulk-renew and
# POST /api/v1/certificates/bulk-reassign) that the GUI calls
# in a single round-trip. Pre-L-1 the GUI looped per-cert
# HTTP calls — 100 selected certs = 100 round-trips × ~50200ms
# each = a 520-second wedge during which the operator stares
# at a progress bar.
#
# This step grep-fails the build if either loop shape reappears
# in CertificatesPage.tsx. Patterns catch the actual pre-L-1
# shapes:
# - `for (const id of ids) { await triggerRenewal(id) }`
# - `for (const id of ids) { await updateCertificate(id, { owner_id }) }`
# - `for (let i = 0; i < ids.length; i++) { await triggerRenewal(ids[i]) }`
#
# Allowed: comment lines explaining the pre-L-1 pattern in the
# docblock above each handler. Test files (_test.tsx) exempt
# so negative-pattern tests can keep working.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-l-fa0c1ac07ab5 and cat-l-8a1fb258a38a for closure
# rationale, or web/src/api/client.ts::bulkRenewCertificates
# / bulkReassignCertificates for the canonical call path.
run: |
set -e
BAD_LOOP=$(grep -nE 'for[[:space:]]*\(' web/src/pages/CertificatesPage.tsx 2>/dev/null \
| grep -E 'await[[:space:]]+(triggerRenewal|updateCertificate)\(' \
| grep -v '\.test\.' \
| grep -vE '^\s*[^:]+:[0-9]+:\s*//' \
|| true)
if [ -n "$BAD_LOOP" ]; then
echo "L-1 regression: client-side bulk-action loop reappeared in CertificatesPage.tsx:"
echo "$BAD_LOOP"
echo ""
echo "Use bulkRenewCertificates({ certificate_ids: [...] }) or"
echo "bulkReassignCertificates({ certificate_ids: [...], owner_id, team_id? })"
echo "instead of looping per-item HTTP calls. See"
echo "coverage-gap-audit-2026-04-24-v5/unified-audit.md cat-l-* for rationale."
exit 1
fi
- name: Forbidden orphan-CRUD client function regression guard (B-1)
# B-1 master closed four audit findings — three orphan-update fns
# (cat-b-31ceb6aaa9f1, cat-b-7a34f893a8f9) and one orphan CRUD
# surface (cat-b-4631ca092bee, RenewalPolicy) — by wiring per-page
# Edit modals so every backend write endpoint has at least one
# GUI consumer. The fourth finding (cat-b-9b97ffb35ef7) deleted
# the dead `exportCertificatePEM` duplicate.
#
# Pre-B-1 the failure mode was: backend ships a CRUD handler,
# client.ts ships the matching `update*` / `delete*` / `create*`
# function, but no page imports it. Operators were forced to
# `psql` directly to edit team names, owner emails, agent-group
# match rules, issuer names, profile names, or any renewal-policy
# field — turning a 30-second GUI task into a 30-minute database
# excursion with audit-trail gaps.
#
# This step fails the build if any of the eight previously-orphan
# client functions loses its page consumer (i.e. a future refactor
# accidentally re-orphans them). Each fn must have ≥1 non-test
# consumer under web/src/pages/. Tests (*.test.ts(x)) and the
# client.ts definition file itself are exempt.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-b-31ceb6aaa9f1, cat-b-7a34f893a8f9, cat-b-4631ca092bee,
# cat-b-9b97ffb35ef7 for closure rationale.
run: |
set -e
ORPHAN_FNS="updateOwner updateTeam updateAgentGroup updateIssuer updateProfile createRenewalPolicy updateRenewalPolicy deleteRenewalPolicy"
FAIL=0
for fn in $ORPHAN_FNS; do
HITS=$(grep -rE "\b${fn}\b" web/src/pages/ 2>/dev/null \
| grep -vE '\.test\.(ts|tsx):' \
| wc -l)
if [ "$HITS" -eq 0 ]; then
echo "::error::B-1 regression: client function '${fn}' has zero consumers under web/src/pages/."
echo " Every backend CRUD endpoint must have a GUI consumer to avoid forcing operators to psql."
echo " Either restore the page consumer or delete the client function in the same commit."
FAIL=1
fi
done
# cat-b-9b97ffb35ef7: exportCertificatePEM was deleted as a dead
# duplicate of downloadCertificatePEM. Block resurrection.
if grep -nE 'export\s+const\s+exportCertificatePEM' web/src/api/client.ts >/dev/null 2>&1; then
echo "::error::B-1 regression: exportCertificatePEM was removed as a dead duplicate of downloadCertificatePEM."
echo " If a JSON variant is needed, add an explicit page consumer in the same commit."
FAIL=1
fi
if [ "$FAIL" -ne 0 ]; then
exit 1
fi
echo "B-1 orphan-CRUD client function guardrail: all 8 functions have page consumers."
- name: Forbidden strings.Contains(err.Error()) regression guard (S-2)
# S-2 closure (cat-s6-efc7f6f6bd50): replaced 30 brittle
# substring-match error-dispatch sites in internal/api/handler/
# with errors.Is + typed sentinels (repository.ErrNotFound,
# repository.ErrForeignKeyConstraint via the
# repository.IsForeignKeyError helper). This step grep-fails
# the build if any new strings.Contains(err.Error(), "not found")
# or strings.Contains(err.Error(), "violates foreign key")
# site appears under internal/api/handler/.
#
# Allowed: closure-comments documenting the convention (e.g.
# bulk_reassignment.go's "post-M-1 errToStatus convention"
# docblock); domain-specific substring patterns that are
# legitimately one-off ("cannot approve", "cannot reject",
# "cannot be parsed", "challenge password") — flagged as
# deferred follow-ups in the S-2 commit message.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-s6-efc7f6f6bd50 for closure rationale.
run: |
set -e
BAD=$(grep -rnE 'strings\.Contains\(err\.Error\(\),\s*"(not found|violates foreign key|RESTRICT)"' internal/api/handler/ 2>/dev/null \
| grep -vE '^\s*[^:]+:[0-9]+:\s*//' \
|| true)
if [ -n "$BAD" ]; then
echo "S-2 regression: brittle substring-match error-dispatch reappeared:"
echo "$BAD"
echo ""
echo "Use errors.Is(err, repository.ErrNotFound) for not-found dispatch,"
echo "or repository.IsForeignKeyError(err) for FK violations."
echo "See coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-s6-efc7f6f6bd50 for closure rationale."
exit 1
fi
echo "S-2 typed-sentinel error-dispatch guardrail: clean."
- name: Race Detection
run: go test -race ./internal/service/... ./internal/api/handler/... ./internal/api/middleware/... ./internal/scheduler/... ./internal/connector/... ./internal/crypto/... ./internal/domain/... ./internal/validation/... ./internal/tlsprobe/... -count=1 -timeout 300s
- name: Go Test with Coverage
run: |
go test ./internal/service/... ./internal/api/handler/... ./internal/api/middleware/... ./internal/integration/... ./internal/connector/issuer/... ./internal/connector/target/... ./internal/connector/notifier/... ./internal/connector/discovery/... ./internal/crypto/... ./internal/mcp/... ./internal/cli/... ./internal/domain/... ./internal/validation/... ./internal/tlsprobe/... -count=1 -cover -coverprofile=coverage.out
- name: Check Coverage Thresholds
run: |
# Extract per-package coverage from test output
echo "=== Coverage Report ==="
go tool cover -func=coverage.out | tail -1
# Check service layer coverage (target: 60%+)
SERVICE_COV=$(go tool cover -func=coverage.out | grep 'internal/service' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Service layer coverage: ${SERVICE_COV}%"
# Check handler layer coverage (target: 60%+)
HANDLER_COV=$(go tool cover -func=coverage.out | grep 'internal/api/handler' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Handler layer coverage: ${HANDLER_COV}%"
# Check domain layer coverage (target: 40%+)
DOMAIN_COV=$(go tool cover -func=coverage.out | grep 'internal/domain' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Domain layer coverage: ${DOMAIN_COV}%"
# Check middleware layer coverage (target: 50%+)
MIDDLEWARE_COV=$(go tool cover -func=coverage.out | grep 'internal/api/middleware' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Middleware layer coverage: ${MIDDLEWARE_COV}%"
# Check crypto package coverage (target: 85%+)
# M-8 rationale: encryption primitives are a security-critical gate.
# v2 format, key-derivation, fallback, and fail-closed sentinel paths
# all need exhaustive coverage to avoid silent regressions (CWE-916 / CWE-329).
CRYPTO_COV=$(go tool cover -func=coverage.out | grep 'internal/crypto' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Crypto package coverage: ${CRYPTO_COV}%"
# Bundle-7 / Audit H-005 — extended crypto-cluster gates per CLAUDE.md.
# internal/pkcs7/ is at 100% at HEAD (encoder-only, exhaustively tested
# via Bundle-4 fuzz targets + unit tests). internal/connector/issuer/local/
# is at 68.3% at HEAD; H-010 tracks the gap and will lift this floor
# to 85% once the missing CSR-validation + CA-cert-loading tests land.
PKCS7_COV=$(go tool cover -func=coverage.out | grep 'internal/pkcs7' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "PKCS7 package coverage: ${PKCS7_COV}%"
LOCAL_ISSUER_COV=$(go tool cover -func=coverage.out | grep 'internal/connector/issuer/local' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "Local-issuer coverage: ${LOCAL_ISSUER_COV}%"
# Bundle-J / Coverage-Audit C-001 (partial-closed) — ACME failure-mode
# batch lifts internal/connector/issuer/acme from 41.8% to ~55.6%
# (per-package package-scoped run). The global per-file average can
# come in lower because this awk pattern divides by file count
# rather than weighting by line count, but with the failure-mode
# tests landed every file in the package has at least 50% coverage.
# Floor set at 50 to accommodate the global-run arithmetic; bumps
# to 85 when Bundle J-extended (Pebble-style mock) lands and the
# IssueCertificate / solveAuthorizations* flows are exercisable.
ACME_COV=$(go tool cover -func=coverage.out | grep 'internal/connector/issuer/acme' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "ACME issuer coverage: ${ACME_COV}%"
# Bundle-L.B / Coverage-Audit C-005 — StepCA failure-mode + JWE
# round-trip tests lift internal/connector/issuer/stepca from
# 52.1% to 90.4% (per-package run). Floor at 80 with margin.
STEPCA_COV=$(go tool cover -func=coverage.out | grep 'internal/connector/issuer/stepca' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "StepCA issuer coverage: ${STEPCA_COV}%"
# Bundle-K / Coverage-Audit C-002 — MCP per-tool dispatch via
# in-memory transport lifts internal/mcp from 28.0% to 93.1%
# (per-package run). Floor at 85.
MCP_COV=$(go tool cover -func=coverage.out | grep 'internal/mcp/' | awk '{print $NF}' | sed 's/%//' | awk '{sum+=$1; n++} END {if(n>0) printf "%.1f", sum/n; else print "0"}')
echo "MCP coverage: ${MCP_COV}%"
# Fail if thresholds not met.
# Bundle R-CI-extended raises (post-Bundle-N.C-extended):
# service 55 -> 70 (HEAD 73.4%; 3pp margin); handler 60 -> 75
# (HEAD 79.8%; 4pp margin). Prescribed Bundle R target was 80;
# held lower to avoid false-positives on single low-coverage
# files dragging the global per-file-average down.
if [ "$(echo "$SERVICE_COV < 70" | bc -l)" -eq 1 ]; then
echo "::error::Service layer coverage ${SERVICE_COV}% is below 70% (Bundle R-CI-extended floor — add tests, do not lower the gate)"
exit 1
fi
if [ "$(echo "$HANDLER_COV < 75" | bc -l)" -eq 1 ]; then
echo "::error::Handler layer coverage ${HANDLER_COV}% is below 75% (Bundle R-CI-extended floor — add tests, do not lower the gate)"
exit 1
fi
if [ "$(echo "$DOMAIN_COV < 40" | bc -l)" -eq 1 ]; then
echo "::error::Domain layer coverage ${DOMAIN_COV}% is below 40% threshold"
exit 1
fi
if [ "$(echo "$MIDDLEWARE_COV < 30" | bc -l)" -eq 1 ]; then
echo "::error::Middleware layer coverage ${MIDDLEWARE_COV}% is below 30% threshold"
exit 1
fi
# Bundle R / Coverage Audit Closure — CI threshold raise checkpoint #3.
# Crypto package floor lifted 85 → 88. Post-Bundle-Q package-scoped
# coverage at HEAD: 88.2% (Bundle Q's gopter property tests don't add
# production-code coverage — they exercise the same paths via
# generative inputs). The remaining ~12% gap is platform-failure
# branches (rand.Reader / aes.NewCipher) that require interface seams
# the production code doesn't use; closing them is tracked as
# R-CI-extended, not Bundle R scope.
if [ "$(echo "$CRYPTO_COV < 88" | bc -l)" -eq 1 ]; then
echo "::error::Crypto package coverage ${CRYPTO_COV}% is below 88% (Bundle R closure floor — add tests, do not lower the gate)"
exit 1
fi
# Bundle-7 / H-005: pkcs7 coverage is INFORMATIONAL only in this run.
# The global `go test -cover ./...` invocation in CI doesn't exercise
# internal/pkcs7's tests (they're primarily Fuzz* targets that
# require an explicit `-fuzz` invocation, plus encoder helpers
# exercised transitively). The deep-scan workflow runs
# `go test -cover ./internal/pkcs7/...` directly and confirmed 100%
# at Bundle-7 close — that's the load-bearing measurement. Keeping
# the global-run number visible here for trend-watching but not
# gating because 0% is a measurement artifact, not a regression.
echo "PKCS7 package coverage (global run, informational): ${PKCS7_COV}%"
# Bundle-9 / H-010 closure: local-issuer HARD gate at 85%. The
# transitional 60% floor (Bundle-7) was an explicit promise in the
# CI config that H-010 would raise it once CSR-validation + CA-
# cert-loading + key-rotation + key-encoding pin tests landed.
# Bundle-9 ships those tests (bundle9_coverage_test.go) and lifts
# the package-scoped run to ~86.7%; the global run averages a few
# points lower (per-function arithmetic), so the gate is set to 85
# with the live `go test -cover` number being the source of truth.
# If this gate trips, the fix is to add tests, NOT to lower the
# floor — every percentage point under 85 is a regression on the
# H-010 closure invariant.
# Bundle R / Coverage Audit Closure — CI threshold raise checkpoint #3.
# Local-issuer floor lifted 85 → 86. Post-Bundle-Q package-scoped
# coverage at HEAD: 86.7%. The prescribed Bundle R target was
# 92, but reaching it requires interface seams for crypto/x509
# signing-error branches — tracked as R-CI-extended.
if [ "$(echo "$LOCAL_ISSUER_COV < 86" | bc -l)" -eq 1 ]; then
echo "::error::Local-issuer coverage ${LOCAL_ISSUER_COV}% is below 86% (Bundle R closure floor — add tests, do not lower the gate)"
exit 1
fi
# Bundle R-CI-extended threshold raise (post-Bundle-J-extended):
# ACME 50 -> 80. The Pebble-style mock + per-CA failure tests
# lift package-scoped ACME to 85.4%; gate at 80 with 5pp margin
# to absorb the global-run per-file-average dip. The prescribed
# Bundle R target was 85; held at 80 to avoid false-positives
# on single low-coverage files.
if [ "$(echo "$ACME_COV < 80" | bc -l)" -eq 1 ]; then
echo "::error::ACME issuer coverage ${ACME_COV}% is below 80% (Bundle R-CI-extended floor — add tests, do not lower the gate)"
exit 1
fi
if [ "$(echo "$STEPCA_COV < 80" | bc -l)" -eq 1 ]; then
echo "::error::StepCA issuer coverage ${STEPCA_COV}% is below 80% (Bundle L.B closure floor — add tests, do not lower the gate)"
exit 1
fi
if [ "$(echo "$MCP_COV < 85" | bc -l)" -eq 1 ]; then
echo "::error::MCP coverage ${MCP_COV}% is below 85% (Bundle K closure floor — add tests, do not lower the gate)"
exit 1
fi
echo "Coverage thresholds passed!"
- name: Upload Coverage Report
uses: actions/upload-artifact@v4
with:
name: go-coverage
path: coverage.out
retention-days: 30
# Bundle P / Strengthening #6 — QA-doc drift guards. Forces every PR
# that adds a Part to docs/testing-guide.md OR a seed row to
# migrations/seed_demo.sql to keep docs/qa-test-guide.md in sync. This
# eliminates the doc-drift class structurally — the symptom Bundle I
# had to clean up by hand becomes a CI-time error going forward.
- name: QA-doc Part-count drift guard
run: |
set -e
DOC_PARTS=$(grep -oE '49 of [0-9]+ Parts' docs/qa-test-guide.md | grep -oE '[0-9]+' | tail -1)
GUIDE_PARTS=$(grep -cE '^## Part [0-9]+:' docs/testing-guide.md)
if [ -z "$DOC_PARTS" ]; then
echo "::error::Could not extract Part count from docs/qa-test-guide.md headline."
echo " Expected pattern: '49 of <N> Parts'"
exit 1
fi
if [ "$DOC_PARTS" != "$GUIDE_PARTS" ]; then
echo "::error::DRIFT — qa-test-guide.md headline claims $DOC_PARTS Parts; testing-guide.md has $GUIDE_PARTS Parts."
echo " Update docs/qa-test-guide.md to match. Bundle I patched this once;"
echo " Bundle P added this guard so the drift cannot recur silently."
exit 1
fi
echo "QA-doc Part-count drift guard: clean ($DOC_PARTS == $GUIDE_PARTS)."
- name: QA-doc seed-count drift guard
run: |
set -e
# Seed-cert count: agnostic to documented header format. The current
# documented count lives in `### Certificates (32 total in ...` —
# extract the first integer in that header.
DOC_CERTS=$(grep -oE '### Certificates \([0-9]+' docs/qa-test-guide.md | grep -oE '[0-9]+' | head -1)
# Authoritative count: unique mc-* IDs in seed_demo.sql.
SEED_CERTS=$(grep -oE 'mc-[a-z0-9_-]+' migrations/seed_demo.sql | sort -u | wc -l | tr -d ' ')
if [ -z "$DOC_CERTS" ]; then
echo "::warning::Could not extract documented cert count from docs/qa-test-guide.md."
echo " Skipping cert-count drift check (header format may have changed)."
elif [ "$DOC_CERTS" != "$SEED_CERTS" ]; then
echo "::error::DRIFT — qa-test-guide.md says $DOC_CERTS certs; seed_demo.sql has $SEED_CERTS unique mc-* IDs."
echo " Update docs/qa-test-guide.md::Seed Data Reference to match."
exit 1
fi
# Issuers: seed-table count vs doc claim.
DOC_ISS=$(grep -oE '### Issuers \([0-9]+' docs/qa-test-guide.md | grep -oE '[0-9]+' | head -1)
# Authoritative: unique iss-* IDs (close enough proxy; the issuers
# table count IS the unique-ID count for this prefix).
SEED_ISS=$(grep -oE 'iss-[a-z0-9_-]+' migrations/seed_demo.sql | sort -u | wc -l | tr -d ' ')
if [ -z "$DOC_ISS" ]; then
echo "::warning::Could not extract documented issuer count."
elif [ "$DOC_ISS" != "$SEED_ISS" ] && [ "$((SEED_ISS - DOC_ISS))" -gt 5 ]; then
# Allow up to 5pp slack — iss-* IDs appear in audit_events and
# other reference tables that aren't issuer-table rows. Drift
# only flags when the spread grows large.
echo "::error::DRIFT — qa-test-guide.md says $DOC_ISS issuers; seed_demo.sql has $SEED_ISS unique iss-* IDs (spread > 5)."
exit 1
fi
echo "QA-doc seed-count drift guard: clean."
# Bundle Q / I-001 closure — test-naming convention guard (informational).
# The convention is `Test<Func>_<Scenario>_<ExpectedResult>`. This step
# prints any non-conformant tests but does NOT fail the build until the
# Bundle I-001-extended (2026-04-27) — promoted from informational
# to hard-fail. The convention is now: every `func TestXxx(...)` MUST
# match Go's standard test-runner pattern (`^func Test[A-Z]`). Tests
# whose name starts with `func Test<lowercase>` are silently SKIPPED
# by `go test` (Go only runs `Test[A-Z]...`) — those are the real
# bugs this guard catches.
#
# The original audit's `Test<Func>_<Scenario>_<ExpectedResult>` triple-
# token prescription has been relaxed: single-function pin tests like
# `TestNewAgent` or `TestSplitPEMChain` are valid Go convention, with
# internal scenarios expressed via `t.Run` subtests. Requiring the
# underscore-Scenario-Result triple repo-wide would mean renaming
# 167 legitimate tests for no observable behavior change. The
# Test<Func>_<Scenario>_<ExpectedResult> form remains documented as
# the recommended pattern for parameterized scenarios in
# docs/qa-test-guide.md, but is not gated.
- name: Test-naming convention guard (hard-fail)
run: |
# Catch tests Go itself would silently skip: `func TestX...` where
# the first letter after `Test` is lowercase. Go's testing runner
# requires uppercase to register the test; lowercase tests don't
# run, which is a real bug a CI guard should catch.
INVALID=$(grep -rnE '^func Test[a-z]' --include='*_test.go' . \
| grep -v '_test.go.bak' \
|| true)
if [ -n "$INVALID" ]; then
echo "::error::Found tests Go would silently skip (lowercase after 'Test'):"
echo "$INVALID"
echo "Rename to start with an uppercase letter — Go's test runner only matches ^Test[A-Z]."
exit 1
fi
echo "Test-naming convention guard: clean (no Go-invalid test names found)."
frontend-build:
name: Frontend Build
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: '22'
- name: Install Dependencies
working-directory: web
run: npm ci
- name: TypeScript Check
working-directory: web
run: npx tsc --noEmit
- name: Run Frontend Tests
working-directory: web
run: npx vitest run
- name: Build Frontend
working-directory: web
run: npx vite build
- name: Forbidden hardcoded source-count prose regression guard (S-1)
# S-1 master closed cat-s1-9ce1cbe26876 (README + features.md
# stale numeric counts; explicit CLAUDE.md violation per
# "version-stamped numbers rot") and
# cat-s1-features_md_issuer_count_contradiction (features.md
# self-disagreed on issuer count: 9 vs 12 in the same doc).
# The fix replaced source-derived numbers in prose with
# "rebuild via <command>" patterns documented in CLAUDE.md::
# "Current-state commands". This step grep-fails the build if
# any of the previously-stale sites reintroduces a hardcoded
# count.
#
# Allowed surfaces: demo-fixture prose in README ("32
# certificates" — those are seed_demo.sql facts, not live
# source counts), historical-milestone counts in
# WORKSPACE-CHANGELOG.md, the testing-guide example phrasing
# ("README claims 8 issuer connectors but only 6 exist"),
# and any number that quotes the source command immediately
# adjacent.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-s1-9ce1cbe26876 + cat-s1-features_md_issuer_count_contradiction
# for closure rationale.
run: |
set -e
BAD=$(grep -rnE '\b[0-9]+\s+(issuer connectors?|target connectors?|notifier connectors?|discovery connectors?|MCP tools|OpenAPI operations|migrations|database tables|frontend pages|HTTP routes)\b' \
README.md docs/ 2>/dev/null \
| grep -vE 'WORKSPACE-CHANGELOG|seed_demo|demo override' \
| grep -vE 'DRIFT HAZARD|Source: |Rebuild|rebuild via|grep -|wc -l|ls -d|find ' \
| grep -vE 'README claims [0-9]+ issuer connectors but only [0-9]+ exist' \
|| true)
if [ -n "$BAD" ]; then
echo "S-1 regression: hardcoded source-count prose reappeared:"
echo "$BAD"
echo ""
echo "CLAUDE.md rule: 'Numeric claims about current state rot.'"
echo "Replace the count with the grep command from CLAUDE.md::"
echo "'Current-state commands' (e.g. 'ls -d internal/connector/issuer/*/ | wc -l')"
echo "or rephrase to reference the rebuild command on the same line."
echo "See coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "cat-s1-9ce1cbe26876 for closure rationale."
exit 1
fi
echo "S-1 stale-counts guardrail: clean."
- name: Documented orphan client fns sync guard (P-1)
# P-1 master closed diff-04x03-d24864996ad4 + cat-b-dc46aadab98e
# by documenting 17 detail-page-candidate orphan client.ts
# functions in a docblock at the top of web/src/api/client.ts.
# This step verifies the docblock list ↔ export list relationship:
# every name listed in the docblock must still be declared as
# an export below it (catches drift where someone deletes the
# export but forgets the docblock, or vice versa).
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# diff-04x03-d24864996ad4 + cat-b-dc46aadab98e for closure rationale.
run: |
set -e
# CRL/OCSP-Responder Phase 5 closed the getOCSPStatus orphan: the
# CertificateDetailPage Revocation Endpoints panel now consumes it
# ("Check OCSP status" button). Removed from the list to keep the
# docblock + guardrail honest.
DOCUMENTED='getAgentGroup getAgentGroupMembers getAuditEvent getCertificateDeployments getDiscoveredCertificate getHealthCheck getHealthCheckHistory getNetworkScanTarget getNotification getOwner getPolicy getPolicyViolations getRenewalPolicy getTeam registerAgent updateHealthCheck'
MISSING=""
for fn in $DOCUMENTED; do
if ! grep -qE "^export const ${fn}\b" web/src/api/client.ts; then
MISSING="${MISSING}${fn} "
fi
done
if [ -n "$MISSING" ]; then
echo "P-1 regression: documented orphan(s) missing from client.ts exports:"
echo " $MISSING"
echo ""
echo "Either restore the export, or delete the corresponding line"
echo "in the documented-orphans docblock at the top of client.ts."
echo "See coverage-gap-audit-2026-04-24-v5/unified-audit.md"
echo "diff-04x03-d24864996ad4 for closure rationale."
exit 1
fi
echo "P-1 documented-orphans sync guard: clean ($(echo $DOCUMENTED | wc -w) fns verified)."
- name: Frontend page-coverage regression guard (T-1)
# T-1 closure (cat-s2-c24a548076c6): pre-T-1 only 3 of 28 pages
# had Vitest coverage. T-1 lifted that to 11/28 by writing tests
# for the 8 highest-leverage pages (CertificatesPage filter +
# pagination state, the new B-1 Edit modals, the D-2 type-trim
# render sites, etc.). The remaining pages are deferred to per-
# page commits — when the next feature change touches them, the
# test gets added in the same commit. This step blocks new
# pages from landing without tests.
#
# Allowlist: pages that are explicitly deferred — listed below
# with a one-line "why deferred" justification. Each entry must
# be removed when the page gets its test.
# - LoginPage: static auth form, no business logic
# - AuditPage: read-only timeline; D-2 already trimmed
# - ShortLivedPage: derived view of certs already covered by CertificatesPage
# - DigestPage: server-rendered digest; minimal client logic
# - ObservabilityPage: exposes Prometheus / Grafana links only
# - HealthMonitorPage: wraps M-006 health check timeline; M-006 has its own tests
# - NetworkScanPage: wraps the network scanner UX; SSRF unit-tested in domain
# - JobsPage: covered transitively via AgentDetailPage
# - JobDetailPage: drill-down view; covered transitively via JobsPage
# - AgentFleetPage: bulk overview; covered transitively via AgentsPage
# - ProfilesPage: CRUD form; mirrors PoliciesPage shape (covered)
# - CertificateDetailPage: drill-down view; covered transitively via CertificatesPage
# - IssuerDetailPage: drill-down view; covered transitively via IssuersPage
# - TargetDetailPage: drill-down view; covered transitively via TargetsPage
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-s2-c24a548076c6 for closure rationale.
run: |
set -e
ALLOW='^(LoginPage|AuditPage|ShortLivedPage|DigestPage|ObservabilityPage|HealthMonitorPage|NetworkScanPage|JobsPage|JobDetailPage|AgentFleetPage|ProfilesPage|CertificateDetailPage|IssuerDetailPage|TargetDetailPage)$'
UNTESTED=""
for f in web/src/pages/*.tsx; do
base=$(basename "$f" .tsx)
case "$f" in *.test.tsx) continue ;; esac
if [ -f "web/src/pages/${base}.test.tsx" ]; then continue; fi
if echo "$base" | grep -qE "$ALLOW"; then continue; fi
UNTESTED="${UNTESTED}${base} "
done
if [ -n "$UNTESTED" ]; then
echo "T-1 regression: page(s) without sibling .test.tsx and not on the deferred allowlist:"
echo " $UNTESTED"
echo ""
echo "Either add web/src/pages/<Page>.test.tsx (mirror NotificationsPage.test.tsx),"
echo "or add the page to the ALLOW pattern in .github/workflows/ci.yml with a"
echo "one-line 'why deferred' comment. See"
echo "coverage-gap-audit-2026-04-24-v5/unified-audit.md cat-s2-c24a548076c6"
echo "for closure rationale."
exit 1
fi
ALLOWLIST_SIZE=$(echo "$ALLOW" | tr '|' '\n' | wc -l)
echo "T-1 page-coverage guardrail: clean (allowlist size: $ALLOWLIST_SIZE pages deferred)."
- name: Bundle-8 / L-015 target=_blank rel=noopener regression guard
# Audit L-015 / CWE-1022 (reverse-tabnabbing): every <a target="_blank">
# MUST carry rel="noopener noreferrer" so a malicious page at the
# target URL cannot navigate the opener window via window.opener.
# At Bundle-8 close (commit b566355→) all 3 sites in the codebase
# already comply — this guard prevents regression. The
# ExternalLink component (web/src/components/ExternalLink.tsx)
# is the recommended way to add new external links.
#
# Test files (web/src/**/*.test.{ts,tsx}) are excluded so test
# docstrings or fixture data describing the attack vector by
# name don't trip the guard — symmetric with the L-019 guard.
run: |
set -e
OFFENDERS=$(grep -rnE 'target=["'"'"']?_blank["'"'"']?' web/src/ 2>/dev/null \
| grep -v 'noopener noreferrer' \
| grep -v 'web/src/components/ExternalLink.tsx' \
| grep -vE '\.test\.(ts|tsx)(:[0-9]+)?:' \
|| true)
if [ -n "$OFFENDERS" ]; then
echo "L-015 regression: target=\"_blank\" without rel=\"noopener noreferrer\":"
echo "$OFFENDERS"
echo ""
echo "Either add rel=\"noopener noreferrer\" inline,"
echo "or migrate to <ExternalLink> from web/src/components/ExternalLink.tsx."
exit 1
fi
echo "L-015 target=_blank guardrail: clean."
- name: Bundle-8 / L-019 dangerouslySetInnerHTML regression guard
# Audit L-019 / CWE-79 (XSS): no PRODUCTION code may use
# dangerouslySetInnerHTML directly. At Bundle-8 close the codebase
# has 0 sites; future genuine needs MUST route through
# web/src/utils/safeHtml.ts::sanitizeHtml.
#
# Test files (web/src/**/*.test.{ts,tsx}) are explicitly excluded:
# the M-029 Pass 3 XSS-hardening test docstrings legitimately cite
# the attack vector by name to explain what the test is guarding
# against (e.g. "a careless refactor to dangerouslySetInnerHTML
# would let an attacker-controlled CSR deliver an XSS payload").
# Tests describing the threat aren't using it; the guard's intent
# is production code only.
run: |
set -e
OFFENDERS=$(grep -rnE 'dangerouslySetInnerHTML' web/src/ 2>/dev/null \
| grep -v 'web/src/utils/safeHtml.ts' \
| grep -vE '\.test\.(ts|tsx)(:[0-9]+)?:' \
|| true)
if [ -n "$OFFENDERS" ]; then
echo "L-019 regression: dangerouslySetInnerHTML used outside safeHtml.ts:"
echo "$OFFENDERS"
echo ""
echo "Route through web/src/utils/safeHtml.ts::sanitizeHtml — see file"
echo "header for the activation procedure (DOMPurify dependency)."
exit 1
fi
echo "L-019 dangerouslySetInnerHTML guardrail: clean."
- name: Bundle-8 / M-009 + M-029 Pass 1 mutation contract guard (hard zero)
# Audit M-009 + M-029 Pass 1 closure:
#
# Pre-Bundle-8 the codebase had 56 bare useMutation sites with
# discretionary invalidation. Bundle 8 shipped the useTrackedMutation
# wrapper (web/src/hooks/useTrackedMutation.ts) that requires every
# caller to declare `invalidates: QueryKey[] | 'noop'`. M-029 Pass 1
# then migrated all 56 sites to the wrapper across 6 batches.
#
# This guard pins the contract going forward: every useMutation call
# in src/ MUST be inside useTrackedMutation.ts (the wrapper itself
# is the only legitimate caller of useMutation). Any bare useMutation
# call elsewhere is a regression — adding a new mutation site means
# going through the wrapper so the invalidates contract is enforced
# per-site, not by a soft budget guard.
#
# If you genuinely need raw useMutation (extremely unlikely — the
# wrapper supports invalidates: 'noop' for fire-and-forget mutations),
# update this guard's exclusion list and document the carve-out.
run: |
set -e
# Test files (web/src/**/*.test.{ts,tsx}) are excluded so existing
# useMutation-mocking test patterns and the wrapper's own unit
# tests don't trip the production guard — symmetric with L-015
# and L-019 above.
BARE=$(grep -rnE '\buseMutation\(' web/src/ 2>/dev/null \
| grep -v 'web/src/hooks/useTrackedMutation\.ts' \
| grep -vE '\.test\.(ts|tsx)(:[0-9]+)?:' \
|| true)
if [ -n "$BARE" ]; then
echo "M-009 hard-zero regression: bare useMutation() call(s) outside the wrapper:"
echo "$BARE"
echo
echo "Every mutation must go through useTrackedMutation"
echo "(web/src/hooks/useTrackedMutation.ts) with explicit"
echo "invalidates: QueryKey[] | 'noop'. See file header for usage."
exit 1
fi
# Sanity counts (informational, not a gate).
TRACKED=$(grep -rcE '\buseTrackedMutation\(' web/src/ 2>/dev/null | awk -F: '{s+=$2} END{print s}')
INVALIDATIONS=$(grep -rcE 'invalidateQueries|setQueryData|removeQueries|invalidates:' web/src/ 2>/dev/null | awk -F: '{s+=$2} END{print s}')
echo "M-009 hard-zero: bare useMutation sites = 0 (wrapper-internal call + test files excluded)."
echo "M-009 informational: useTrackedMutation sites = $TRACKED; invalidation surface = $INVALIDATIONS."
- name: Forbidden env-var docs drift regression guard (G-3)
# G-3 master closed cat-g-163dae19bc59 (docs-only env vars
# phantom in features.md), cat-g-b8f8f8796159 (6 config-only
# env vars never documented), and cat-g-renewal_check_interval_rename_drift
# (features.md still advertised the pre-rename
# CERTCTL_RENEWAL_CHECK_INTERVAL after it was renamed to
# CERTCTL_SCHEDULER_RENEWAL_CHECK_INTERVAL). This step runs
# `comm -23` both ways between the env vars defined in Go
# source (config.go + cmd/agent + deploy/test fixtures + ACME
# DNS-01 script env exports) and the env vars mentioned in
# README + docs/ + deploy/helm/.
#
# Allowlist: env vars that are documented as integration-
# surface contracts (script env exports for ACME DNS-01,
# OpenSSL CA scripts, StepCA per-issuer-config-blob fields,
# Webhook per-notifier-config-blob fields, ACME EAB, audit
# exclusion, demo-stack overrides) but not consumed directly
# by config.go. Each entry below has a one-line justification
# — if you add a new entry, add the justification too.
#
# See coverage-gap-audit-2026-04-24-v5/unified-audit.md
# cat-g-* for closure rationale.
run: |
set -e
# Defined: config.go + agent + cli + mcp-server + server cmds + test fixtures + ACME DNS export
{
grep -nE '"CERTCTL_[A-Z_]+"' internal/config/config.go | sed -E 's/.*"(CERTCTL_[A-Z_]+)".*/\1/'
grep -rhoE '"CERTCTL_[A-Z_]+"' cmd/agent/*.go cmd/cli/*.go cmd/mcp-server/*.go cmd/server/*.go 2>/dev/null | sed -E 's/"(CERTCTL_[A-Z_]+)"/\1/'
grep -rhoE 'CERTCTL_[A-Z_]+' deploy/test/qa_test.go internal/connector/issuer/acme/dns.go 2>/dev/null
} | grep -E '^CERTCTL_' | sort -u > /tmp/g3-defined.txt
# Documented: README + docs + helm
grep -rhoE '\bCERTCTL_[A-Z_]+\b' README.md docs/ deploy/helm/ 2>/dev/null | sort -u > /tmp/g3-docs.txt
# Allowlist of env vars documented as external integration contracts.
# Each entry justifies itself in one line; if you add to this list,
# add the justification.
ALLOWED='^(
CERTCTL_OPENSSL_SIGN_SCRIPT|
CERTCTL_OPENSSL_REVOKE_SCRIPT|
CERTCTL_OPENSSL_CRL_SCRIPT|
CERTCTL_OPENSSL_TIMEOUT_SECONDS|
CERTCTL_STEPCA_URL|
CERTCTL_STEPCA_FINGERPRINT|
CERTCTL_STEPCA_PROVISIONER|
CERTCTL_STEPCA_PROVISIONER_NAME|
CERTCTL_STEPCA_PROVISIONER_KEY|
CERTCTL_STEPCA_PROVISIONER_JWK|
CERTCTL_STEPCA_PROVISIONER_PASSWORD|
CERTCTL_STEPCA_PASSWORD|
CERTCTL_STEPCA_KEY_PATH|
CERTCTL_STEPCA_ROOT_CA|
CERTCTL_WEBHOOK_URL|
CERTCTL_WEBHOOK_SECRET|
CERTCTL_ACME_EAB_KID|
CERTCTL_ACME_EAB_HMAC|
CERTCTL_ACME_DNS_PROPAGATION_WAIT|
CERTCTL_AUDIT_EXCLUDE_PATHS|
CERTCTL_TLS_|
CERTCTL_TLS_INSECURE_SKIP_VERIFY|
CERTCTL_SCEP_|
CERTCTL_SERVER_CA_BUNDLE_PATH|
CERTCTL_SERVER_TLS_INSECURE_SKIP_VERIFY|
CERTCTL_QA_[A-Z_]+
)$'
# ^ The CERTCTL_OPENSSL_* / CERTCTL_STEPCA_* / CERTCTL_WEBHOOK_* /
# CERTCTL_ACME_EAB_* / CERTCTL_ACME_DNS_PROPAGATION_WAIT /
# CERTCTL_AUDIT_EXCLUDE_PATHS / CERTCTL_TLS_* / CERTCTL_SERVER_* /
# CERTCTL_QA_* sets are documented integration-surface contracts
# (script invocations, per-issuer config-blob field names,
# per-notifier config-blob field names, demo-stack overrides,
# test fixtures) — not server-side env vars in config.go.
# The audit's "37 docs-only" count over-flagged these; the
# closure narrows the gate to the specific drift sites
# (renewal-interval rename + 6 config-only) and allowlists
# the documented external contracts here.
ALLOWED_FLAT=$(echo "$ALLOWED" | tr -d '\n ')
DOCS_ONLY=$(comm -13 /tmp/g3-defined.txt /tmp/g3-docs.txt | grep -vE "$ALLOWED_FLAT" || true)
CONFIG_ONLY=$(comm -23 /tmp/g3-defined.txt /tmp/g3-docs.txt || true)
if [ -n "$DOCS_ONLY" ]; then
echo "G-3 regression: env var(s) mentioned in docs but not defined in Go source AND not in the documented integration-surface allowlist:"
echo "$DOCS_ONLY"
echo ""
echo "Either delete from docs (phantom/typo) or add to config.go,"
echo "or add to the ALLOWED list with a one-line justification."
exit 1
fi
if [ -n "$CONFIG_ONLY" ]; then
echo "G-3 regression: env var(s) defined in Go source but never documented:"
echo "$CONFIG_ONLY"
echo ""
echo "Add an entry to docs/features.md (or another canonical doc) so operators can find it."
exit 1
fi
echo "G-3 env-var docs drift guardrail: clean."
helm-lint:
name: Helm Chart Validation
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install Helm
uses: azure/setup-helm@v4
with:
version: '3.13.0'
# HTTPS-Everywhere (v2.0.47): the chart fails render when no TLS source is
# configured. Every lint/template invocation below must pick exactly one
# provisioning mode — see deploy/helm/certctl/templates/_helpers.tpl
# (certctl.tls.required) and docs/tls.md.
- name: Lint Helm Chart
run: |
helm lint deploy/helm/certctl/ \
--set server.tls.existingSecret=certctl-tls-ci
- name: Template Helm Chart (existingSecret mode)
run: |
helm template certctl deploy/helm/certctl/ \
--set server.tls.existingSecret=certctl-tls-ci \
> /dev/null
- name: Template Helm Chart (cert-manager mode)
run: |
helm template certctl deploy/helm/certctl/ \
--set server.tls.certManager.enabled=true \
--set server.tls.certManager.issuerRef.name=letsencrypt-prod \
> /dev/null
- name: Template Helm Chart (guard fails without TLS)
run: |
# Inverse test: the chart MUST refuse to render when no TLS source is
# configured. If this ever renders successfully, the fail-loud guard
# in certctl.tls.required has regressed.
if helm template certctl deploy/helm/certctl/ > /dev/null 2>&1; then
echo "::error::Helm chart rendered without a TLS source — fail-loud guard regressed"
exit 1
fi