From e3c65689bbe949e1434e1e2a535023f43b4a76ba Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 09:49:57 -0500
Subject: [PATCH 01/36] copilot: first draft

---
 README.md                      |  3 ++
 check-project-links/README.md  | 40 +++++++++++++++++
 check-project-links/action.yml | 78 ++++++++++++++++++++++++++++++++++
 3 files changed, 121 insertions(+)
 create mode 100644 check-project-links/README.md
 create mode 100644 check-project-links/action.yml

diff --git a/README.md b/README.md
index b134f3b..f760978 100644
--- a/README.md
+++ b/README.md
@@ -13,6 +13,9 @@
 - [`ni/python-actions/check-project-version`](check-project-version): uses Poetry to get the version
   of a Python project and checks that it matches an expected version. Publish workflows can use this
   to verify that the release tag matches the version number in `pyproject.toml`.
+- [`ni/python-actions/check-project-links`](check-project-links): scans project `pyproject.toml` files
+  for URLs, writes the discovered links to a temporary file, and validates each URL in Docker while
+  failing only on `4xx` responses.
 - [`ni/python-actions/update-project-version`](update-project-version): uses Poetry to update the
   version of a Python project and creates a pull request to modify its `pyproject.toml` file.
   Publish workflows can use this to update the version in `pyproject.toml` for the next build.
diff --git a/check-project-links/README.md b/check-project-links/README.md
new file mode 100644
index 0000000..404b425
--- /dev/null
+++ b/check-project-links/README.md
@@ -0,0 +1,40 @@
+# `ni/python-actions/check-project-links`
+
+This action searches a project tree for `pyproject.toml` files, extracts `https://` links from those files, and validates each discovered URL by making an HTTP request in a Docker container. It fails only when a checked URL returns a `4xx` response. `2xx` and `3xx` responses are treated as successful and logged without failing the action.
+
+## Inputs
+
+### `project-directory`
+
+Path to the directory containing one or more `pyproject.toml` files.
+
+Default: `${{ github.workspace }}`
+
+### `docker-image`
+
+Docker image used to perform the HTTP requests.
+
+Default: `curlimages/curl:8.22.0`
+
+## Example
+
+```yaml
+steps:
+  - uses: actions/checkout@v4
+
+  - name: Check project links
+    uses: ni/python-actions/check-project-links@v1
+    with:
+      project-directory: .
+      docker-image: curlimages/curl:8.22.0
+```
+
+## Behavior
+
+- Uses `rg` to locate `pyproject.toml` files beneath the configured project directory.
+- Extracts every `https://...` URL that ends at the first whitespace character.
+- Deduplicates the list of URLs and writes them to a temporary file.
+- Validates each URL with the configured Docker image.
+- Logs `2xx` and `3xx` responses as passing.
+- Fails the action only when a URL returns a `4xx` status code.
+- Other status codes are considered a warning.
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
new file mode 100644
index 0000000..6c8dd5a
--- /dev/null
+++ b/check-project-links/action.yml
@@ -0,0 +1,78 @@
+name: Check project links
+description: Find URLs referenced in pyproject.toml files and fail only when a response is 4xx.
+
+inputs:
+  project-directory:
+    description: Path to the directory containing pyproject.toml files.
+    default: ${{ github.workspace }}
+  docker-image:
+    description: Docker image used to validate each discovered URL.
+    default: curlimages/curl:8.22.0@sha256:58adaa4e8dca9c988bae2aba4ab3434a0bb2da16bbe3f92dec39ec7785166777
+
+runs:
+  using: composite
+  steps:
+    - name: Check project links
+      shell: bash
+      run: |
+        set -u
+
+        project_directory="${{ inputs.project-directory }}"
+        docker_image="${{ inputs.docker-image }}"
+
+        if [ ! -d "$project_directory" ]; then
+          echo "::error title=Check Project Links Error::Project directory '$project_directory' does not exist."
+          exit 1
+        fi
+
+        link_file="$(mktemp)"
+        cleanup() {
+          rm -f "$link_file"
+        }
+        trap cleanup EXIT
+
+        while IFS= read -r manifest; do
+          if [ -n "$manifest" ]; then
+            rg -o 'https://[^[:space:]]+' "$manifest" || true
+            rg 'http://[^[:space:]]+' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring linke." || true
+          fi
+        done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$link_file"
+
+        if [ ! -s "$link_file" ]; then
+          echo "No project links found under $project_directory."
+          exit 0
+        fi
+
+        echo "Found $(wc -l < "$link_file") unique links:"
+        cat "$link_file"
+
+        docker run --rm \
+          -v "$link_file:/tmp/project_links.txt:ro" \
+          "$docker_image" \
+          sh -ec '
+            failed=0
+            while IFS= read -r url; do
+              [ -n "$url" ] || continue
+              status=$(curl -L -sS -o /tmp/link_body -w "%{http_code}" "$url" || true)
+              case "$status" in
+                2??|3??)
+                  echo "PASS $url -> $status"
+                  ;;
+                4??)
+                  echo "FAIL $url -> $status"
+                  failed=1
+                  ;;
+                *)
+                  echo "WARN $url -> $status"
+                  ;;
+              esac
+            done < /tmp/project_links.txt
+            exit "$failed"
+          '
+
+        status=$?
+        if [ "$status" -ne 0 ]; then
+          exit 1
+        fi
+
+        echo "Success: No checked project links returned 4xx responses."

From ab0adce0540554fcf23660d1b2cd0dc3e52b7878 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 09:58:33 -0500
Subject: [PATCH 02/36] copilot: setup tests

---
 .github/workflows/test_actions.yml | 58 ++++++++++++++++++++++++++++++
 1 file changed, 58 insertions(+)

diff --git a/.github/workflows/test_actions.yml b/.github/workflows/test_actions.yml
index 8e1f79a..52e4a18 100644
--- a/.github/workflows/test_actions.yml
+++ b/.github/workflows/test_actions.yml
@@ -254,6 +254,63 @@ jobs:
           project-directory: test-project
           expected-version: 1.0.2.dev1
 
+  test_check_project_links:
+    name: Test check-project-links
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repo
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        with:
+          persist-credentials: false
+      - name: Create project with a valid link
+        run: |
+          mkdir -p test-project
+          cat > test-project/pyproject.toml <<'EOF'
+          [project]
+          name = "test-project"
+          version = "0.1.0"
+          description = "Test project with a valid URL"
+          authors = [{ name = "NI" }]
+          readme = "README.md"
+          requires-python = ">=3.9"
+
+          [project.urls]
+          Homepage = "https://github.com/ni/python-actions"
+          Documentation = "https://github.com/ni/python-actions/blob/main/README.md"
+          EOF
+        shell: bash
+      - name: Check valid project links
+        uses: ./check-project-links
+        with:
+          project-directory: test-project
+      - name: Create project with a missing link
+        run: |
+          mkdir -p failing-project
+          cat > failing-project/pyproject.toml <<'EOF'
+          [project]
+          name = "failing-project"
+          version = "0.1.0"
+          description = "Test project with a 404 URL"
+          authors = [{ name = "NI" }]
+          readme = "README.md"
+          requires-python = ">=3.9"
+
+          [project.urls]
+          Broken = "https://github.com/ni/project-that-does-not-exist"
+          EOF
+        shell: bash
+      - name: Check missing project link (expected to fail)
+        id: expected-failure
+        continue-on-error: true
+        uses: ./check-project-links
+        with:
+          project-directory: failing-project
+      - name: Error if the previous step didn't fail
+        if: steps.expected-failure.outcome != 'failure'
+        run: |
+          echo "::error title=Test Failure::The previous step did not fail as expected."
+          exit 1
+
   test_analyze_project:
     name: Test analyze-project
     runs-on: ${{ matrix.os }}
@@ -454,6 +511,7 @@ jobs:
       test_setup_poetry_no_cache,
       test_check_project_version,
       test_update_project_version,
+      test_check_project_links,
       test_analyze_project,
       test_analyze_project_repo_root,
     ]

From 705a350f36b6de31673883567251d8325258b8a5 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 09:58:41 -0500
Subject: [PATCH 03/36] fix zizmor issues

---
 check-project-links/action.yml | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 6c8dd5a..f9d4b4e 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -14,11 +14,14 @@ runs:
   steps:
     - name: Check project links
       shell: bash
+      env:
+        PROJECT_DIRECTORY: ${{ inputs.project-directory }}
+        DOCKER_IMAGE: ${{ inputs.docker-image }}
       run: |
         set -u
 
-        project_directory="${{ inputs.project-directory }}"
-        docker_image="${{ inputs.docker-image }}"
+        project_directory="$PROJECT_DIRECTORY"
+        docker_image="$DOCKER_IMAGE"
 
         if [ ! -d "$project_directory" ]; then
           echo "::error title=Check Project Links Error::Project directory '$project_directory' does not exist."

From 30bef4ffd877699f2423adc544faa43df7cdba0c Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 09:58:54 -0500
Subject: [PATCH 04/36] use @v0 in doc and add disclaimer

---
 check-project-links/README.md | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index 404b425..c1516ec 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -16,11 +16,16 @@ Docker image used to perform the HTTP requests.
 
 Default: `curlimages/curl:8.22.0`
 
-## Example
+## Examples
+
+> [!NOTE]
+> These examples use `@v0`, but pinning to a commit hash or full release tag is recommended for
+> build reproducibility and security.
+
 
 ```yaml
 steps:
-  - uses: actions/checkout@v4
+  - uses: actions/checkout@v0
 
   - name: Check project links
     uses: ni/python-actions/check-project-links@v1

From 6116fb7285471ad8bff2f24267740191274fbf75 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:01:13 -0500
Subject: [PATCH 05/36] don't care about the full link for http warning

---
 check-project-links/action.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index f9d4b4e..7f57c7e 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -37,7 +37,7 @@ runs:
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
             rg -o 'https://[^[:space:]]+' "$manifest" || true
-            rg 'http://[^[:space:]]+' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring linke." || true
+            rg 'http://' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring linke." || true
           fi
         done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$link_file"
 

From fd83fa7ea2e25c87689e8198ad976b5aba893dba Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:02:57 -0500
Subject: [PATCH 06/36] also handle quote ended links

---
 check-project-links/action.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 7f57c7e..7ea2f91 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -36,7 +36,7 @@ runs:
 
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
-            rg -o 'https://[^[:space:]]+' "$manifest" || true
+            rg -o "https://[^\"[:space:]]+" "$manifest" || true
             rg 'http://' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring linke." || true
           fi
         done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$link_file"

From 44028acfd944a2f45e7e589f7ccf5f3ba81a7748 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:09:52 -0500
Subject: [PATCH 07/36] fix copilot suggested issues

---
 check-project-links/README.md  | 3 +++
 check-project-links/action.yml | 2 +-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index c1516ec..0df0229 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -16,6 +16,9 @@ Docker image used to perform the HTTP requests.
 
 Default: `curlimages/curl:8.22.0`
 
+> [!NOTE]
+> The action default uses a full digest sha. Although the action does not require this.
+
 ## Examples
 
 > [!NOTE]
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 7ea2f91..9b92cd6 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -37,7 +37,7 @@ runs:
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
             rg -o "https://[^\"[:space:]]+" "$manifest" || true
-            rg 'http://' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring linke." || true
+            rg 'http://' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link." || true
           fi
         done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$link_file"
 

From 882ba28d2480a737159bd3507c9de57899cc55a5 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:24:59 -0500
Subject: [PATCH 08/36] add trusted domains input and limit to only checking
 https links at those domains

---
 check-project-links/action.yml | 80 +++++++++++++++++++++++++++++++---
 1 file changed, 74 insertions(+), 6 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 9b92cd6..d2832c7 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -5,6 +5,9 @@ inputs:
   project-directory:
     description: Path to the directory containing pyproject.toml files.
     default: ${{ github.workspace }}
+  allowed-domains:
+    description: Comma-separated list of trusted hostnames or domains to validate. Supports wildcards like *.readthedocs.io.
+    default: github.com,ni.github.io,*.readthedocs.io
   docker-image:
     description: Docker image used to validate each discovered URL.
     default: curlimages/curl:8.22.0@sha256:58adaa4e8dca9c988bae2aba4ab3434a0bb2da16bbe3f92dec39ec7785166777
@@ -16,11 +19,13 @@ runs:
       shell: bash
       env:
         PROJECT_DIRECTORY: ${{ inputs.project-directory }}
+        ALLOWED_DOMAINS: ${{ inputs.allowed-domains }}
         DOCKER_IMAGE: ${{ inputs.docker-image }}
       run: |
         set -u
 
         project_directory="$PROJECT_DIRECTORY"
+        allowed_domains="$ALLOWED_DOMAINS"
         docker_image="$DOCKER_IMAGE"
 
         if [ ! -d "$project_directory" ]; then
@@ -28,25 +33,88 @@ runs:
           exit 1
         fi
 
+        manifest_links="$(mktemp)"
         link_file="$(mktemp)"
         cleanup() {
-          rm -f "$link_file"
+          rm -f "$manifest_links" "$link_file"
         }
         trap cleanup EXIT
 
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
-            rg -o "https://[^\"[:space:]]+" "$manifest" || true
-            rg 'http://' "$manifest" && echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link." || true
+            rg -o "https\?://[^\"[:space:]]+" "$manifest" || true
+            if rg 'http://' "$manifest" >/dev/null; then
+              echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
+            fi
           fi
-        done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$link_file"
+        done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$manifest_links"
+
+        : > "$link_file"
+        while IFS= read -r url; do
+          [ -n "$url" ] || continue
+
+          if printf '%s\n' "$url" | grep -q 'http://'; then
+            echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $url. Ignoring link."
+            continue
+          fi
+
+          host="${url#*://}"
+          host="${host%%/*}"
+          host="${host%%:*}"
+          host="${host,,}"
+          host="${host%.}"
+
+          if printf '%s\n' "$host" | grep -Eq '^(localhost|.*\.localhost)$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^(127\.|0\.0\.0\.0|::1|169\.254\.|metadata(\.google\.internal)?)$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^(10\.|192\.168\.|172\.(1[6-9]|2[0-9]|3[0-1])\.|169\.254\.|fc[0-9a-fA-F:]+|fe[89ab][0-9a-fA-F:]+)$'; then
+            continue
+          fi
+
+          allowed=0
+          IFS=',' read -ra domains <<< "$allowed_domains"
+          for domain in "${domains[@]}"; do
+            domain="${domain//[[:space:]]/}"
+            [ -n "$domain" ] || continue
+            domain="${domain,,}"
+            domain="${domain%.}"
+
+            if [ "${domain:0:2}" = '*.' ]; then
+              suffix="${domain#*.}"
+              if printf '%s\n' "$host" | grep -Eq "^([A-Za-z0-9-]+\.)*${suffix//./\\.}$"; then
+                allowed=1
+                break
+              fi
+            else
+              if [ "$host" = "$domain" ] || printf '%s\n' "$host" | grep -Eq "^([A-Za-z0-9-]+\.)+${domain//./\\.}$"; then
+                allowed=1
+                break
+              fi
+            fi
+          done
+
+          if [ "$allowed" -eq 1 ]; then
+            printf '%s\n' "$url" >> "$link_file"
+          fi
+        done < "$manifest_links"
+
+        # Explicitly drop insecure HTTP links before validation.
+        grep -v 'http://' "$link_file" > "${link_file}.https" || true
+        if [ -s "${link_file}.https" ]; then
+          mv "${link_file}.https" "$link_file"
+        else
+          : > "$link_file"
+        fi
 
         if [ ! -s "$link_file" ]; then
-          echo "No project links found under $project_directory."
+          echo "No trusted project links found under $project_directory."
           exit 0
         fi
 
-        echo "Found $(wc -l < "$link_file") unique links:"
+        echo "Found $(wc -l < "$link_file") unique trusted links:"
         cat "$link_file"
 
         docker run --rm \

From c04646dbb925fec6b4752ca6b0f3d0bda70966cd Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:32:57 -0500
Subject: [PATCH 09/36] add checks for rg and docker commands

---
 check-project-links/README.md  |  4 +++-
 check-project-links/action.yml | 34 ++++++++++++++++++++++++++++++----
 2 files changed, 33 insertions(+), 5 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index 0df0229..d5cce48 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -39,10 +39,12 @@ steps:
 
 ## Behavior
 
-- Uses `rg` to locate `pyproject.toml` files beneath the configured project directory.
+- Uses `rg` to locate `pyproject.toml` files beneath the configured project directory when available.
+- Falls back to `grep` with a warning if `rg` is not installed; consider adding a pre-step to install `rg` for faster runtime.
 - Extracts every `https://...` URL that ends at the first whitespace character.
 - Deduplicates the list of URLs and writes them to a temporary file.
 - Validates each URL with the configured Docker image.
+- Fails immediately if `docker` is not installed or available on `PATH`.
 - Logs `2xx` and `3xx` responses as passing.
 - Fails the action only when a URL returns a `4xx` status code.
 - Other status codes are considered a warning.
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index d2832c7..cfa972e 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -24,6 +24,12 @@ runs:
       run: |
         set -u
 
+
+        if ! command -v docker >/dev/null 2>&1; then
+          echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
+          exit 1
+        fi
+
         project_directory="$PROJECT_DIRECTORY"
         allowed_domains="$ALLOWED_DOMAINS"
         docker_image="$DOCKER_IMAGE"
@@ -40,14 +46,34 @@ runs:
         }
         trap cleanup EXIT
 
+        if command -v rg >/dev/null 2>&1; then
+          rg_available=1
+        else
+          rg_available=0
+          echo "::warning title=Check Project Links Warning::rg is not available; falling back to grep. Consider adding a pre-step to install rg for faster runtime."
+        fi
+
+        if [ "$rg_available" -eq 1 ]; then
+          manifest_list_cmd="rg --files -g 'pyproject.toml' \"$project_directory\""
+        else
+          manifest_list_cmd="find \"$project_directory\" -type f -name 'pyproject.toml'"
+        fi
+
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
-            rg -o "https\?://[^\"[:space:]]+" "$manifest" || true
-            if rg 'http://' "$manifest" >/dev/null; then
-              echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
+            if [ "$rg_available" -eq 1 ]; then
+              rg -o "https\?://[^\"[:space:]]+" "$manifest" || true
+              if rg 'http://' "$manifest" >/dev/null 2>&1; then
+                echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
+              fi
+            else
+              grep -Eo 'https?://[^[:space:]\"]+' "$manifest" || true
+              if grep -Eq 'http://' "$manifest"; then
+                echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
+              fi
             fi
           fi
-        done < <(rg --files -g 'pyproject.toml' "$project_directory") | sort -u > "$manifest_links"
+        done < <(eval "$manifest_list_cmd") | sort -u > "$manifest_links"
 
         : > "$link_file"
         while IFS= read -r url; do

From e3d97a18cbb57e060419e232a4b1f8d894e820a6 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:33:12 -0500
Subject: [PATCH 10/36] whitespace

---
 check-project-links/action.yml | 1 -
 1 file changed, 1 deletion(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index cfa972e..9931fa6 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -24,7 +24,6 @@ runs:
       run: |
         set -u
 
-
         if ! command -v docker >/dev/null 2>&1; then
           echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
           exit 1

From 35617bd76e6da820b5798198ebe3e4c0be4cd0c5 Mon Sep 17 00:00:00 2001
From: mshafer-NI <23644905+mshafer-NI@users.noreply.github.com>
Date: Wed, 2 Sep 2026 10:43:33 -0500
Subject: [PATCH 11/36] Apply batched suggestions from code review

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
---
 check-project-links/README.md  | 11 ++++++++---
 check-project-links/action.yml | 24 +++++++++++-------------
 2 files changed, 19 insertions(+), 16 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index d5cce48..d0aff4d 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -10,15 +10,20 @@ Path to the directory containing one or more `pyproject.toml` files.
 
 Default: `${{ github.workspace }}`
 
+### `allowed-domains`
+
+Comma-separated list of trusted hostnames or domains to validate. Supports wildcards like `*.readthedocs.io`.
+
+Default: `github.com,ni.github.io,*.readthedocs.io`
+
 ### `docker-image`
 
 Docker image used to perform the HTTP requests.
 
-Default: `curlimages/curl:8.22.0`
+Default: `curlimages/curl:8.22.0@sha256:58adaa4e8dca9c988bae2aba4ab3434a0bb2da16bbe3f92dec39ec7785166777`
 
 > [!NOTE]
-> The action default uses a full digest sha. Although the action does not require this.
-
+> The action default uses a full digest SHA, though this is not required.
 ## Examples
 
 > [!NOTE]
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index cfa972e..eed8a2a 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -42,7 +42,7 @@ runs:
         manifest_links="$(mktemp)"
         link_file="$(mktemp)"
         cleanup() {
-          rm -f "$manifest_links" "$link_file"
+          rm -f "$manifest_links" "$link_file" "${link_file}.https"
         }
         trap cleanup EXIT
 
@@ -54,9 +54,9 @@ runs:
         fi
 
         if [ "$rg_available" -eq 1 ]; then
-          manifest_list_cmd="rg --files -g 'pyproject.toml' \"$project_directory\""
+          manifest_list_cmd=(rg --files -g 'pyproject.toml' "$project_directory")
         else
-          manifest_list_cmd="find \"$project_directory\" -type f -name 'pyproject.toml'"
+          manifest_list_cmd=(find "$project_directory" -type f -name 'pyproject.toml')
         fi
 
         while IFS= read -r manifest; do
@@ -73,7 +73,7 @@ runs:
               fi
             fi
           fi
-        done < <(eval "$manifest_list_cmd") | sort -u > "$manifest_links"
+        done < <("${manifest_list_cmd[@]}") | sort -u > "$manifest_links"
 
         : > "$link_file"
         while IFS= read -r url; do
@@ -110,15 +110,13 @@ runs:
 
             if [ "${domain:0:2}" = '*.' ]; then
               suffix="${domain#*.}"
-              if printf '%s\n' "$host" | grep -Eq "^([A-Za-z0-9-]+\.)*${suffix//./\\.}$"; then
-                allowed=1
-                break
-              fi
+              case "$host" in
+                "$suffix"|*."$suffix") allowed=1; break ;;
+              esac
             else
-              if [ "$host" = "$domain" ] || printf '%s\n' "$host" | grep -Eq "^([A-Za-z0-9-]+\.)+${domain//./\\.}$"; then
-                allowed=1
-                break
-              fi
+              case "$host" in
+                "$domain"|*."$domain") allowed=1; break ;;
+              esac
             fi
           done
 
@@ -150,7 +148,7 @@ runs:
             failed=0
             while IFS= read -r url; do
               [ -n "$url" ] || continue
-              status=$(curl -L -sS -o /tmp/link_body -w "%{http_code}" "$url" || true)
+              status=$(curl -L -sS --connect-timeout 5 --max-time 20 -o /tmp/link_body -w "%{http_code}" "$url" || true)
               case "$status" in
                 2??|3??)
                   echo "PASS $url -> $status"

From c7b1ff9230da31ddd17fc8f1d946f990eea65cd8 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 10:50:50 -0500
Subject: [PATCH 12/36] explicitly drop loopback and IP based hosts

---
 check-project-links/README.md  |  1 +
 check-project-links/action.yml | 38 +++++++++++++++++++++++++++++++++-
 2 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index d0aff4d..464252b 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -47,6 +47,7 @@ steps:
 - Uses `rg` to locate `pyproject.toml` files beneath the configured project directory when available.
 - Falls back to `grep` with a warning if `rg` is not installed; consider adding a pre-step to install `rg` for faster runtime.
 - Extracts every `https://...` URL that ends at the first whitespace character.
+- Drops any URL whose hostname is `localhost`, a local loopback address, or any literal IP address before validation.
 - Deduplicates the list of URLs and writes them to a temporary file.
 - Validates each URL with the configured Docker image.
 - Fails immediately if `docker` is not installed or available on `PATH`.
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index c2e7b8a..43761a0 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -41,7 +41,7 @@ runs:
         manifest_links="$(mktemp)"
         link_file="$(mktemp)"
         cleanup() {
-          rm -f "$manifest_links" "$link_file" "${link_file}.https"
+          rm -f "$manifest_links" "$link_file" "${link_file}.https" "${link_file}.filtered"
         }
         trap cleanup EXIT
 
@@ -132,6 +132,42 @@ runs:
           : > "$link_file"
         fi
 
+        # Explicitly drop localhost and any IP-based hostname before validation.
+        : > "${link_file}.filtered"
+        while IFS= read -r url; do
+          [ -n "$url" ] || continue
+
+          host="${url#*://}"
+          host="${host%%/*}"
+          host="${host%%:*}"
+          host="${host,,}"
+          host="${host%.}"
+
+          if printf '%s\n' "$host" | grep -Eq '^(localhost|.*\.localhost)$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^[0-9]+(\.[0-9]+){3}$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^(::1|[0-9a-fA-F:]+)$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^(127\.|0\.0\.0\.0|169\.254\.|metadata(\.google\.internal)?)$'; then
+            continue
+          fi
+          if printf '%s\n' "$host" | grep -Eq '^(10\.|192\.168\.|172\.(1[6-9]|2[0-9]|3[0-1])\.|fc[0-9a-fA-F:]+|fe[89ab][0-9a-fA-F:]+)$'; then
+            continue
+          fi
+
+          printf '%s\n' "$url" >> "${link_file}.filtered"
+        done < "$link_file"
+
+        if [ -s "${link_file}.filtered" ]; then
+          mv "${link_file}.filtered" "$link_file"
+        else
+          : > "$link_file"
+        fi
+
         if [ ! -s "$link_file" ]; then
           echo "No trusted project links found under $project_directory."
           exit 0

From bd73f6a99653f36a2450fcc90ecbf0cec497cc05 Mon Sep 17 00:00:00 2001
From: mshafer-NI <23644905+mshafer-NI@users.noreply.github.com>
Date: Wed, 2 Sep 2026 11:56:44 -0500
Subject: [PATCH 13/36] Apply batched suggestions from code review

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
---
 check-project-links/README.md  | 4 ++--
 check-project-links/action.yml | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index 464252b..8544d59 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -45,8 +45,8 @@ steps:
 ## Behavior
 
 - Uses `rg` to locate `pyproject.toml` files beneath the configured project directory when available.
-- Falls back to `grep` with a warning if `rg` is not installed; consider adding a pre-step to install `rg` for faster runtime.
-- Extracts every `https://...` URL that ends at the first whitespace character.
+- Falls back to `find` (discovery) and `grep` (extraction) with a warning if `rg` is not installed; consider adding a pre-step to install `rg` for faster runtime.
+- Extracts every `https://...` URL that ends at the first whitespace or double-quote character.
 - Drops any URL whose hostname is `localhost`, a local loopback address, or any literal IP address before validation.
 - Deduplicates the list of URLs and writes them to a temporary file.
 - Validates each URL with the configured Docker image.
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 43761a0..0de2dae 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -61,12 +61,12 @@ runs:
         while IFS= read -r manifest; do
           if [ -n "$manifest" ]; then
             if [ "$rg_available" -eq 1 ]; then
-              rg -o "https\?://[^\"[:space:]]+" "$manifest" || true
+              rg -o 'https://[^"[:space:]]+' "$manifest" || true
               if rg 'http://' "$manifest" >/dev/null 2>&1; then
                 echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
               fi
             else
-              grep -Eo 'https?://[^[:space:]\"]+' "$manifest" || true
+              grep -Eo 'https://[^[:space:]\"]+' "$manifest" || true
               if grep -Eq 'http://' "$manifest"; then
                 echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
               fi

From 8943bd40cafa70b460d1075575cbe250eb2371d4 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 16:24:14 -0500
Subject: [PATCH 14/36] use existing env vars directly

---
 check-project-links/action.yml | 18 +++++++-----------
 1 file changed, 7 insertions(+), 11 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 43761a0..e0a79a4 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -29,12 +29,8 @@ runs:
           exit 1
         fi
 
-        project_directory="$PROJECT_DIRECTORY"
-        allowed_domains="$ALLOWED_DOMAINS"
-        docker_image="$DOCKER_IMAGE"
-
-        if [ ! -d "$project_directory" ]; then
-          echo "::error title=Check Project Links Error::Project directory '$project_directory' does not exist."
+        if [ ! -d "$PROJECT_DIRECTORY" ]; then
+          echo "::error title=Check Project Links Error::Project directory '$PROJECT_DIRECTORY' does not exist."
           exit 1
         fi
 
@@ -53,9 +49,9 @@ runs:
         fi
 
         if [ "$rg_available" -eq 1 ]; then
-          manifest_list_cmd=(rg --files -g 'pyproject.toml' "$project_directory")
+          manifest_list_cmd=(rg --files -g 'pyproject.toml' "$PROJECT_DIRECTORY")
         else
-          manifest_list_cmd=(find "$project_directory" -type f -name 'pyproject.toml')
+          manifest_list_cmd=(find "$PROJECT_DIRECTORY" -type f -name 'pyproject.toml')
         fi
 
         while IFS= read -r manifest; do
@@ -100,7 +96,7 @@ runs:
           fi
 
           allowed=0
-          IFS=',' read -ra domains <<< "$allowed_domains"
+          IFS=',' read -ra domains <<< "$ALLOWED_DOMAINS"
           for domain in "${domains[@]}"; do
             domain="${domain//[[:space:]]/}"
             [ -n "$domain" ] || continue
@@ -169,7 +165,7 @@ runs:
         fi
 
         if [ ! -s "$link_file" ]; then
-          echo "No trusted project links found under $project_directory."
+          echo "No trusted project links found under $PROJECT_DIRECTORY."
           exit 0
         fi
 
@@ -178,7 +174,7 @@ runs:
 
         docker run --rm \
           -v "$link_file:/tmp/project_links.txt:ro" \
-          "$docker_image" \
+          "$DOCKER_IMAGE" \
           sh -ec '
             failed=0
             while IFS= read -r url; do

From b61454c82584755b5a4486713ccde9f6535a1881 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 16:44:06 -0500
Subject: [PATCH 15/36] move it to a py file

---
 check-project-links/action.yml            | 161 +++++-----------------
 check-project-links/find_project_links.py | 102 ++++++++++++++
 2 files changed, 133 insertions(+), 130 deletions(-)
 create mode 100644 check-project-links/find_project_links.py

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index e0a79a4..caaa8de 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -15,165 +15,66 @@ inputs:
 runs:
   using: composite
   steps:
-    - name: Check project links
+    - name: Get project links
+      id: get_project_links
       shell: bash
       env:
         PROJECT_DIRECTORY: ${{ inputs.project-directory }}
         ALLOWED_DOMAINS: ${{ inputs.allowed-domains }}
-        DOCKER_IMAGE: ${{ inputs.docker-image }}
       run: |
         set -u
 
-        if ! command -v docker >/dev/null 2>&1; then
-          echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
-          exit 1
-        fi
-
         if [ ! -d "$PROJECT_DIRECTORY" ]; then
           echo "::error title=Check Project Links Error::Project directory '$PROJECT_DIRECTORY' does not exist."
           exit 1
         fi
 
-        manifest_links="$(mktemp)"
         link_file="$(mktemp)"
         cleanup() {
-          rm -f "$manifest_links" "$link_file" "${link_file}.https" "${link_file}.filtered"
+          rm -f "$link_file"
         }
         trap cleanup EXIT
 
-        if command -v rg >/dev/null 2>&1; then
-          rg_available=1
-        else
-          rg_available=0
-          echo "::warning title=Check Project Links Warning::rg is not available; falling back to grep. Consider adding a pre-step to install rg for faster runtime."
-        fi
+        python3 "$GITHUB_ACTION_PATH/find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
 
-        if [ "$rg_available" -eq 1 ]; then
-          manifest_list_cmd=(rg --files -g 'pyproject.toml' "$PROJECT_DIRECTORY")
-        else
-          manifest_list_cmd=(find "$PROJECT_DIRECTORY" -type f -name 'pyproject.toml')
+        echo "link-file=$link_file" >> "$GITHUB_OUTPUT"
+
+        if [ ! -s "$link_file" ]; then
+          echo "No trusted project links found under $PROJECT_DIRECTORY."
+          exit 0
         fi
 
-        while IFS= read -r manifest; do
-          if [ -n "$manifest" ]; then
-            if [ "$rg_available" -eq 1 ]; then
-              rg -o "https\?://[^\"[:space:]]+" "$manifest" || true
-              if rg 'http://' "$manifest" >/dev/null 2>&1; then
-                echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
-              fi
-            else
-              grep -Eo 'https?://[^[:space:]\"]+' "$manifest" || true
-              if grep -Eq 'http://' "$manifest"; then
-                echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $manifest. Consider using HTTPS instead. Ignoring link."
-              fi
-            fi
-          fi
-        done < <("${manifest_list_cmd[@]}") | sort -u > "$manifest_links"
-
-        : > "$link_file"
-        while IFS= read -r url; do
-          [ -n "$url" ] || continue
-
-          if printf '%s\n' "$url" | grep -q 'http://'; then
-            echo "::warning title=Check Project Links Warning::Found insecure HTTP link in $url. Ignoring link."
-            continue
-          fi
-
-          host="${url#*://}"
-          host="${host%%/*}"
-          host="${host%%:*}"
-          host="${host,,}"
-          host="${host%.}"
-
-          if printf '%s\n' "$host" | grep -Eq '^(localhost|.*\.localhost)$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^(127\.|0\.0\.0\.0|::1|169\.254\.|metadata(\.google\.internal)?)$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^(10\.|192\.168\.|172\.(1[6-9]|2[0-9]|3[0-1])\.|169\.254\.|fc[0-9a-fA-F:]+|fe[89ab][0-9a-fA-F:]+)$'; then
-            continue
-          fi
-
-          allowed=0
-          IFS=',' read -ra domains <<< "$ALLOWED_DOMAINS"
-          for domain in "${domains[@]}"; do
-            domain="${domain//[[:space:]]/}"
-            [ -n "$domain" ] || continue
-            domain="${domain,,}"
-            domain="${domain%.}"
-
-            if [ "${domain:0:2}" = '*.' ]; then
-              suffix="${domain#*.}"
-              case "$host" in
-                "$suffix"|*."$suffix") allowed=1; break ;;
-              esac
-            else
-              case "$host" in
-                "$domain"|*."$domain") allowed=1; break ;;
-              esac
-            fi
-          done
-
-          if [ "$allowed" -eq 1 ]; then
-            printf '%s\n' "$url" >> "$link_file"
-          fi
-        done < "$manifest_links"
-
-        # Explicitly drop insecure HTTP links before validation.
-        grep -v 'http://' "$link_file" > "${link_file}.https" || true
-        if [ -s "${link_file}.https" ]; then
-          mv "${link_file}.https" "$link_file"
-        else
-          : > "$link_file"
+        echo "Found $(wc -l < "$link_file") unique trusted links:"
+        cat "$link_file"
+
+    - name: Check project links
+      shell: bash
+      env:
+        LINK_FILE: ${{ steps.get_project_links.outputs.link-file }}
+        DOCKER_IMAGE: ${{ inputs.docker-image }}
+      run: |
+        set -u
+
+        if ! command -v docker >/dev/null 2>&1; then
+          echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
+          exit 1
         fi
 
-        # Explicitly drop localhost and any IP-based hostname before validation.
-        : > "${link_file}.filtered"
-        while IFS= read -r url; do
-          [ -n "$url" ] || continue
-
-          host="${url#*://}"
-          host="${host%%/*}"
-          host="${host%%:*}"
-          host="${host,,}"
-          host="${host%.}"
-
-          if printf '%s\n' "$host" | grep -Eq '^(localhost|.*\.localhost)$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^[0-9]+(\.[0-9]+){3}$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^(::1|[0-9a-fA-F:]+)$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^(127\.|0\.0\.0\.0|169\.254\.|metadata(\.google\.internal)?)$'; then
-            continue
-          fi
-          if printf '%s\n' "$host" | grep -Eq '^(10\.|192\.168\.|172\.(1[6-9]|2[0-9]|3[0-1])\.|fc[0-9a-fA-F:]+|fe[89ab][0-9a-fA-F:]+)$'; then
-            continue
-          fi
-
-          printf '%s\n' "$url" >> "${link_file}.filtered"
-        done < "$link_file"
-
-        if [ -s "${link_file}.filtered" ]; then
-          mv "${link_file}.filtered" "$link_file"
-        else
-          : > "$link_file"
+        if [ -z "${LINK_FILE:-}" ] || [ ! -f "$LINK_FILE" ]; then
+          echo "::error title=Check Project Links Error::No project links file was generated."
+          exit 1
         fi
 
-        if [ ! -s "$link_file" ]; then
-          echo "No trusted project links found under $PROJECT_DIRECTORY."
+        if [ ! -s "$LINK_FILE" ]; then
+          echo "No trusted project links found."
           exit 0
         fi
 
-        echo "Found $(wc -l < "$link_file") unique trusted links:"
-        cat "$link_file"
+        echo "Found $(wc -l < "$LINK_FILE") unique trusted links:"
+        cat "$LINK_FILE"
 
         docker run --rm \
-          -v "$link_file:/tmp/project_links.txt:ro" \
+          -v "$LINK_FILE:/tmp/project_links.txt:ro" \
           "$DOCKER_IMAGE" \
           sh -ec '
             failed=0
diff --git a/check-project-links/find_project_links.py b/check-project-links/find_project_links.py
new file mode 100644
index 0000000..074a991
--- /dev/null
+++ b/check-project-links/find_project_links.py
@@ -0,0 +1,102 @@
+#!/usr/bin/env python3
+
+import argparse
+import ipaddress
+import os
+import sys
+from urllib.parse import urlsplit
+
+try:
+    import tomllib
+except ModuleNotFoundError:  # pragma: no cover
+    raise Exception('tomllib is not available. Please use Python 3.11 or later.')
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description='Collect trusted project URLs from pyproject.toml files.')
+    parser.add_argument('project_directory', help='Directory containing project pyproject.toml files.')
+    parser.add_argument('allowed_domains', help='Comma-separated list of allowed domains.')
+    parser.add_argument('output_path', help='Path to write discovered URLs.')
+    return parser.parse_args()
+
+
+def is_allowed(hostname: str, allowed: list[str]) -> bool:
+    if not hostname:
+        return False
+
+    host = hostname.lower().rstrip('.')
+    if host == 'localhost' or host.endswith('.localhost'):
+        return False
+
+    try:
+        ip = ipaddress.ip_address(host)
+    except ValueError:
+        ip = None
+
+    if ip is not None:
+        return False
+
+    for domain in allowed:
+        if domain.startswith('*.'):
+            suffix = domain[2:]
+            if host == suffix or host.endswith(f'.{suffix}'):
+                return True
+        else:
+            if host == domain:
+                return True
+    return False
+
+
+def walk_metadata(value, results, allowed):
+    if isinstance(value, dict):
+        for item in value.values():
+            walk_metadata(item, results, allowed)
+    elif isinstance(value, list):
+        for item in value:
+            walk_metadata(item, results, allowed)
+    elif isinstance(value, str):
+        if not value.startswith('https://'):
+            return
+        host = urlsplit(value).hostname
+        if host and is_allowed(host, allowed):
+            results.add(value)
+
+
+def main() -> int:
+    args = parse_args()
+    project_directory = args.project_directory
+    allowed_domains = args.allowed_domains
+    output_path = args.output_path
+
+    allowed = {}
+    for raw_domain in allowed_domains.split(','):
+        domain = raw_domain.strip().lower().rstrip('.')
+        if domain:
+            allowed.add(domain)
+
+    results: set[str] = set()
+
+    for root, _, files in os.walk(project_directory):
+        for file_name in files:
+            if file_name != 'pyproject.toml':
+                continue
+
+            manifest_path = os.path.join(root, file_name)
+            try:
+                with open(manifest_path, 'rb') as manifest_file:
+                    metadata = tomllib.load(manifest_file)
+            except (OSError, tomllib.TOMLDecodeError):
+                print(f"Warning: Failed to read or parse pyproject.toml at {manifest_path}", file=sys.stderr)
+                continue
+
+            walk_metadata(metadata, results, allowed)
+
+    with open(output_path, 'w', encoding='utf-8') as output_file:
+        for url in sorted(results):
+            output_file.write(f'{url}\n')
+
+    return 0
+
+
+if __name__ == '__main__':
+    sys.exit(main())

From a772e336d4c48a75012600eaa3b111dfc373ce72 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:18:12 -0500
Subject: [PATCH 16/36] update doc

---
 check-project-links/README.md | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index 8544d59..bcb4a67 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -44,9 +44,8 @@ steps:
 
 ## Behavior
 
-- Uses `rg` to locate `pyproject.toml` files beneath the configured project directory when available.
-- Falls back to `find` (discovery) and `grep` (extraction) with a warning if `rg` is not installed; consider adding a pre-step to install `rg` for faster runtime.
-- Extracts every `https://...` URL that ends at the first whitespace or double-quote character.
+- Uses Python directory walk and [`tomllib`](https://docs.python.org/3/library/tomllib.html) + string search to identify all links in pyproject.toml files under the provided path.
+- Extracts strings that are parsable URLs, and URLs from comments
 - Drops any URL whose hostname is `localhost`, a local loopback address, or any literal IP address before validation.
 - Deduplicates the list of URLs and writes them to a temporary file.
 - Validates each URL with the configured Docker image.

From 1ef1d05b52e55c8c90bcf4f6fb29af299b8d791b Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:19:09 -0500
Subject: [PATCH 17/36] switch to python impl

---
 check-project-links/action.yml            | 43 +++++--------
 check-project-links/find_project_links.py | 75 ++++++++++++++++++-----
 2 files changed, 75 insertions(+), 43 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index caaa8de..f6ac900 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -15,19 +15,25 @@ inputs:
 runs:
   using: composite
   steps:
-    - name: Get project links
-      id: get_project_links
+    - name: Check project links
+      id: check-project-links
       shell: bash
       env:
         PROJECT_DIRECTORY: ${{ inputs.project-directory }}
         ALLOWED_DOMAINS: ${{ inputs.allowed-domains }}
+        DOCKER_IMAGE: ${{ inputs.docker-image }}
       run: |
-        set -u
+        set -euo pipefail
 
         if [ ! -d "$PROJECT_DIRECTORY" ]; then
           echo "::error title=Check Project Links Error::Project directory '$PROJECT_DIRECTORY' does not exist."
           exit 1
         fi
+        
+        if ! command -v docker >/dev/null 2>&1; then
+          echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
+          exit 1
+        fi
 
         link_file="$(mktemp)"
         cleanup() {
@@ -37,8 +43,6 @@ runs:
 
         python3 "$GITHUB_ACTION_PATH/find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
 
-        echo "link-file=$link_file" >> "$GITHUB_OUTPUT"
-
         if [ ! -s "$link_file" ]; then
           echo "No trusted project links found under $PROJECT_DIRECTORY."
           exit 0
@@ -47,34 +51,17 @@ runs:
         echo "Found $(wc -l < "$link_file") unique trusted links:"
         cat "$link_file"
 
-    - name: Check project links
-      shell: bash
-      env:
-        LINK_FILE: ${{ steps.get_project_links.outputs.link-file }}
-        DOCKER_IMAGE: ${{ inputs.docker-image }}
-      run: |
-        set -u
-
-        if ! command -v docker >/dev/null 2>&1; then
-          echo "::error title=Check Project Links Error::docker is not available. Install Docker or add a pre-step that installs it before using this action."
-          exit 1
-        fi
-
-        if [ -z "${LINK_FILE:-}" ] || [ ! -f "$LINK_FILE" ]; then
-          echo "::error title=Check Project Links Error::No project links file was generated."
-          exit 1
-        fi
-
-        if [ ! -s "$LINK_FILE" ]; then
+        if [ ! -s "$link_file" ]; then
           echo "No trusted project links found."
           exit 0
         fi
 
-        echo "Found $(wc -l < "$LINK_FILE") unique trusted links:"
-        cat "$LINK_FILE"
+        echo "Found $(wc -l < "$link_file") unique trusted links:"
+        cat "$link_file"
 
-        docker run --rm \
-          -v "$LINK_FILE:/tmp/project_links.txt:ro" \
+        unprivileged_user=100:100 # this matches the default, but we want to be explicit about it
+        docker run -u "${unprivileged_user}" --rm \
+          -v "$link_file:/tmp/project_links.txt:ro" \
           "$DOCKER_IMAGE" \
           sh -ec '
             failed=0
diff --git a/check-project-links/find_project_links.py b/check-project-links/find_project_links.py
index 074a991..c969801 100644
--- a/check-project-links/find_project_links.py
+++ b/check-project-links/find_project_links.py
@@ -20,7 +20,19 @@ def parse_args() -> argparse.Namespace:
     return parser.parse_args()
 
 
-def is_allowed(hostname: str, allowed: list[str]) -> bool:
+def is_allowed(hostname: str, allowed: set[str]) -> bool:
+    """
+    Check if the given hostname is allowed based on the provided set of allowed domains.
+
+    >>> is_allowed('example.com', {'example.com'})
+    True
+
+    >>> is_allowed('sub.example.com', {'example.com'})
+    False
+
+    >>> is_allowed('sub.example.com', {'*.example.com'})
+    True
+    """
     if not hostname:
         return False
 
@@ -36,18 +48,30 @@ def is_allowed(hostname: str, allowed: list[str]) -> bool:
     if ip is not None:
         return False
 
-    for domain in allowed:
-        if domain.startswith('*.'):
-            suffix = domain[2:]
-            if host == suffix or host.endswith(f'.{suffix}'):
-                return True
-        else:
-            if host == domain:
-                return True
+    if host in allowed:
+        return True
+
+    if any(host.endswith(f'.{domain.lstrip(".*")}') for domain in allowed if domain.startswith('*.')):
+        return True
+
     return False
 
 
-def walk_metadata(value, results, allowed):
+def safe_under(base_dir: str, candidate_path: str) -> str:
+    """Return a canonical path that stays under base_dir, or raise ValueError."""
+    safe_base = os.path.realpath(base_dir)
+    safe_candidate = os.path.realpath(candidate_path)
+
+    try:
+        if os.path.commonpath([safe_base, safe_candidate]) != safe_base:
+            raise ValueError(f"Path escapes base directory: {candidate_path!r}")
+    except ValueError as exc:
+        raise ValueError(f"Path escapes base directory: {candidate_path!r}") from exc
+
+    return safe_candidate
+
+
+def walk_metadata(value, results: set[str], allowed: set[str]) -> None:
     if isinstance(value, dict):
         for item in value.values():
             walk_metadata(item, results, allowed)
@@ -65,10 +89,12 @@ def walk_metadata(value, results, allowed):
 def main() -> int:
     args = parse_args()
     project_directory = args.project_directory
+    safe_project_directory = os.path.realpath(project_directory, strict=True)
+    safe_project_directory = safe_under(os.getcwd(), safe_project_directory)
     allowed_domains = args.allowed_domains
-    output_path = args.output_path
+    safe_output_path = safe_under("/tmp", os.path.realpath(args.output_path, strict=True))
 
-    allowed = {}
+    allowed: set[str] = set()
     for raw_domain in allowed_domains.split(','):
         domain = raw_domain.strip().lower().rstrip('.')
         if domain:
@@ -76,12 +102,12 @@ def main() -> int:
 
     results: set[str] = set()
 
-    for root, _, files in os.walk(project_directory):
+    for root, _, files in os.walk(safe_project_directory):
         for file_name in files:
             if file_name != 'pyproject.toml':
                 continue
 
-            manifest_path = os.path.join(root, file_name)
+            manifest_path = safe_under(safe_project_directory, os.path.join(root, file_name))
             try:
                 with open(manifest_path, 'rb') as manifest_file:
                     metadata = tomllib.load(manifest_file)
@@ -90,8 +116,27 @@ def main() -> int:
                 continue
 
             walk_metadata(metadata, results, allowed)
+            # Also find any commented URLs in the pyproject.toml file
+            try:
+                with open(manifest_path, 'r', encoding='utf-8') as manifest_file:
+                    for line in manifest_file:
+                        line = line.strip()
+                        prefix, comment = line.split('#', maxsplit=1) if '#' in line else (line, '')
+                        if comment.strip():
+                            comment = comment.strip()
+                            if comment.startswith('https://'):
+                                host = urlsplit(comment).hostname  # handles extra at the end just fine
+                                if host and is_allowed(host, allowed):
+                                    results.add(comment)
+            except OSError:
+                print(f"Warning: Failed to read pyproject.toml at {manifest_path}", file=sys.stderr)
+                continue
+
+    output_directory = os.path.dirname(safe_output_path)
+    if output_directory:
+        os.makedirs(output_directory, exist_ok=True)
 
-    with open(output_path, 'w', encoding='utf-8') as output_file:
+    with open(safe_output_path, 'w', encoding='utf-8') as output_file:
         for url in sorted(results):
             output_file.write(f'{url}\n')
 

From 4c195441daff03b1d053cf1895c0f4c0093fa7ea Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:28:12 -0500
Subject: [PATCH 18/36] all private

---
 check-project-links/find_project_links.py | 36 +++++++++++------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/check-project-links/find_project_links.py b/check-project-links/find_project_links.py
index c969801..1b1203c 100644
--- a/check-project-links/find_project_links.py
+++ b/check-project-links/find_project_links.py
@@ -12,7 +12,7 @@
     raise Exception('tomllib is not available. Please use Python 3.11 or later.')
 
 
-def parse_args() -> argparse.Namespace:
+def _parse_args() -> argparse.Namespace:
     parser = argparse.ArgumentParser(description='Collect trusted project URLs from pyproject.toml files.')
     parser.add_argument('project_directory', help='Directory containing project pyproject.toml files.')
     parser.add_argument('allowed_domains', help='Comma-separated list of allowed domains.')
@@ -20,17 +20,17 @@ def parse_args() -> argparse.Namespace:
     return parser.parse_args()
 
 
-def is_allowed(hostname: str, allowed: set[str]) -> bool:
+def _is_allowed(hostname: str, allowed: set[str]) -> bool:
     """
     Check if the given hostname is allowed based on the provided set of allowed domains.
 
-    >>> is_allowed('example.com', {'example.com'})
+    >>> _is_allowed('example.com', {'example.com'})
     True
 
-    >>> is_allowed('sub.example.com', {'example.com'})
+    >>> _is_allowed('sub.example.com', {'example.com'})
     False
 
-    >>> is_allowed('sub.example.com', {'*.example.com'})
+    >>> _is_allowed('sub.example.com', {'*.example.com'})
     True
     """
     if not hostname:
@@ -57,7 +57,7 @@ def is_allowed(hostname: str, allowed: set[str]) -> bool:
     return False
 
 
-def safe_under(base_dir: str, candidate_path: str) -> str:
+def _safe_under(base_dir: str, candidate_path: str) -> str:
     """Return a canonical path that stays under base_dir, or raise ValueError."""
     safe_base = os.path.realpath(base_dir)
     safe_candidate = os.path.realpath(candidate_path)
@@ -71,28 +71,28 @@ def safe_under(base_dir: str, candidate_path: str) -> str:
     return safe_candidate
 
 
-def walk_metadata(value, results: set[str], allowed: set[str]) -> None:
+def _walk_metadata(value, results: set[str], allowed: set[str]) -> None:
     if isinstance(value, dict):
         for item in value.values():
-            walk_metadata(item, results, allowed)
+            _walk_metadata(item, results, allowed)
     elif isinstance(value, list):
         for item in value:
-            walk_metadata(item, results, allowed)
+            _walk_metadata(item, results, allowed)
     elif isinstance(value, str):
         if not value.startswith('https://'):
             return
         host = urlsplit(value).hostname
-        if host and is_allowed(host, allowed):
+        if host and _is_allowed(host, allowed):
             results.add(value)
 
 
-def main() -> int:
-    args = parse_args()
+def _main() -> int:
+    args = _parse_args()
     project_directory = args.project_directory
     safe_project_directory = os.path.realpath(project_directory, strict=True)
-    safe_project_directory = safe_under(os.getcwd(), safe_project_directory)
+    safe_project_directory = _safe_under(os.getcwd(), safe_project_directory)
     allowed_domains = args.allowed_domains
-    safe_output_path = safe_under("/tmp", os.path.realpath(args.output_path, strict=True))
+    safe_output_path = _safe_under("/tmp", os.path.realpath(args.output_path, strict=True))
 
     allowed: set[str] = set()
     for raw_domain in allowed_domains.split(','):
@@ -107,7 +107,7 @@ def main() -> int:
             if file_name != 'pyproject.toml':
                 continue
 
-            manifest_path = safe_under(safe_project_directory, os.path.join(root, file_name))
+            manifest_path = _safe_under(safe_project_directory, os.path.join(root, file_name))
             try:
                 with open(manifest_path, 'rb') as manifest_file:
                     metadata = tomllib.load(manifest_file)
@@ -115,7 +115,7 @@ def main() -> int:
                 print(f"Warning: Failed to read or parse pyproject.toml at {manifest_path}", file=sys.stderr)
                 continue
 
-            walk_metadata(metadata, results, allowed)
+            _walk_metadata(metadata, results, allowed)
             # Also find any commented URLs in the pyproject.toml file
             try:
                 with open(manifest_path, 'r', encoding='utf-8') as manifest_file:
@@ -126,7 +126,7 @@ def main() -> int:
                             comment = comment.strip()
                             if comment.startswith('https://'):
                                 host = urlsplit(comment).hostname  # handles extra at the end just fine
-                                if host and is_allowed(host, allowed):
+                                if host and _is_allowed(host, allowed):
                                     results.add(comment)
             except OSError:
                 print(f"Warning: Failed to read pyproject.toml at {manifest_path}", file=sys.stderr)
@@ -144,4 +144,4 @@ def main() -> int:
 
 
 if __name__ == '__main__':
-    sys.exit(main())
+    sys.exit(_main())

From d9ead733804df8c5ea6f001903616c9a8bc2854c Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:28:37 -0500
Subject: [PATCH 19/36] fix docstring

---
 check-project-links/find_project_links.py | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/check-project-links/find_project_links.py b/check-project-links/find_project_links.py
index 1b1203c..e8655b0 100644
--- a/check-project-links/find_project_links.py
+++ b/check-project-links/find_project_links.py
@@ -21,8 +21,7 @@ def _parse_args() -> argparse.Namespace:
 
 
 def _is_allowed(hostname: str, allowed: set[str]) -> bool:
-    """
-    Check if the given hostname is allowed based on the provided set of allowed domains.
+    """Check if the given hostname is allowed based on the provided set of allowed domains.
 
     >>> _is_allowed('example.com', {'example.com'})
     True

From d37c307d91a51a95937bd5159469b4ef11b131f4 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:29:38 -0500
Subject: [PATCH 20/36] format

---
 check-project-links/find_project_links.py | 53 ++++++++++++++---------
 1 file changed, 32 insertions(+), 21 deletions(-)

diff --git a/check-project-links/find_project_links.py b/check-project-links/find_project_links.py
index e8655b0..0d9c5ce 100644
--- a/check-project-links/find_project_links.py
+++ b/check-project-links/find_project_links.py
@@ -9,14 +9,18 @@
 try:
     import tomllib
 except ModuleNotFoundError:  # pragma: no cover
-    raise Exception('tomllib is not available. Please use Python 3.11 or later.')
+    raise Exception("tomllib is not available. Please use Python 3.11 or later.")
 
 
 def _parse_args() -> argparse.Namespace:
-    parser = argparse.ArgumentParser(description='Collect trusted project URLs from pyproject.toml files.')
-    parser.add_argument('project_directory', help='Directory containing project pyproject.toml files.')
-    parser.add_argument('allowed_domains', help='Comma-separated list of allowed domains.')
-    parser.add_argument('output_path', help='Path to write discovered URLs.')
+    parser = argparse.ArgumentParser(
+        description="Collect trusted project URLs from pyproject.toml files."
+    )
+    parser.add_argument(
+        "project_directory", help="Directory containing project pyproject.toml files."
+    )
+    parser.add_argument("allowed_domains", help="Comma-separated list of allowed domains.")
+    parser.add_argument("output_path", help="Path to write discovered URLs.")
     return parser.parse_args()
 
 
@@ -35,8 +39,8 @@ def _is_allowed(hostname: str, allowed: set[str]) -> bool:
     if not hostname:
         return False
 
-    host = hostname.lower().rstrip('.')
-    if host == 'localhost' or host.endswith('.localhost'):
+    host = hostname.lower().rstrip(".")
+    if host == "localhost" or host.endswith(".localhost"):
         return False
 
     try:
@@ -50,7 +54,9 @@ def _is_allowed(hostname: str, allowed: set[str]) -> bool:
     if host in allowed:
         return True
 
-    if any(host.endswith(f'.{domain.lstrip(".*")}') for domain in allowed if domain.startswith('*.')):
+    if any(
+        host.endswith(f'.{domain.lstrip(".*")}') for domain in allowed if domain.startswith("*.")
+    ):
         return True
 
     return False
@@ -78,7 +84,7 @@ def _walk_metadata(value, results: set[str], allowed: set[str]) -> None:
         for item in value:
             _walk_metadata(item, results, allowed)
     elif isinstance(value, str):
-        if not value.startswith('https://'):
+        if not value.startswith("https://"):
             return
         host = urlsplit(value).hostname
         if host and _is_allowed(host, allowed):
@@ -94,8 +100,8 @@ def _main() -> int:
     safe_output_path = _safe_under("/tmp", os.path.realpath(args.output_path, strict=True))
 
     allowed: set[str] = set()
-    for raw_domain in allowed_domains.split(','):
-        domain = raw_domain.strip().lower().rstrip('.')
+    for raw_domain in allowed_domains.split(","):
+        domain = raw_domain.strip().lower().rstrip(".")
         if domain:
             allowed.add(domain)
 
@@ -103,28 +109,33 @@ def _main() -> int:
 
     for root, _, files in os.walk(safe_project_directory):
         for file_name in files:
-            if file_name != 'pyproject.toml':
+            if file_name != "pyproject.toml":
                 continue
 
             manifest_path = _safe_under(safe_project_directory, os.path.join(root, file_name))
             try:
-                with open(manifest_path, 'rb') as manifest_file:
+                with open(manifest_path, "rb") as manifest_file:
                     metadata = tomllib.load(manifest_file)
             except (OSError, tomllib.TOMLDecodeError):
-                print(f"Warning: Failed to read or parse pyproject.toml at {manifest_path}", file=sys.stderr)
+                print(
+                    f"Warning: Failed to read or parse pyproject.toml at {manifest_path}",
+                    file=sys.stderr,
+                )
                 continue
 
             _walk_metadata(metadata, results, allowed)
             # Also find any commented URLs in the pyproject.toml file
             try:
-                with open(manifest_path, 'r', encoding='utf-8') as manifest_file:
+                with open(manifest_path, "r", encoding="utf-8") as manifest_file:
                     for line in manifest_file:
                         line = line.strip()
-                        prefix, comment = line.split('#', maxsplit=1) if '#' in line else (line, '')
+                        prefix, comment = line.split("#", maxsplit=1) if "#" in line else (line, "")
                         if comment.strip():
                             comment = comment.strip()
-                            if comment.startswith('https://'):
-                                host = urlsplit(comment).hostname  # handles extra at the end just fine
+                            if comment.startswith("https://"):
+                                host = urlsplit(
+                                    comment
+                                ).hostname  # handles extra at the end just fine
                                 if host and _is_allowed(host, allowed):
                                     results.add(comment)
             except OSError:
@@ -135,12 +146,12 @@ def _main() -> int:
     if output_directory:
         os.makedirs(output_directory, exist_ok=True)
 
-    with open(safe_output_path, 'w', encoding='utf-8') as output_file:
+    with open(safe_output_path, "w", encoding="utf-8") as output_file:
         for url in sorted(results):
-            output_file.write(f'{url}\n')
+            output_file.write(f"{url}\n")
 
     return 0
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     sys.exit(_main())

From ad3ece36623aebbd0047633e528d690738b15565 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:41:48 -0500
Subject: [PATCH 21/36] also not a public module

---
 .../{find_project_links.py => _find_project_links.py}           | 0
 check-project-links/action.yml                                  | 2 +-
 2 files changed, 1 insertion(+), 1 deletion(-)
 rename check-project-links/{find_project_links.py => _find_project_links.py} (100%)

diff --git a/check-project-links/find_project_links.py b/check-project-links/_find_project_links.py
similarity index 100%
rename from check-project-links/find_project_links.py
rename to check-project-links/_find_project_links.py
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index f6ac900..46593dc 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -41,7 +41,7 @@ runs:
         }
         trap cleanup EXIT
 
-        python3 "$GITHUB_ACTION_PATH/find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
+        python3 "$GITHUB_ACTION_PATH/_find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
 
         if [ ! -s "$link_file" ]; then
           echo "No trusted project links found under $PROJECT_DIRECTORY."

From bcbd82feaf83e3556022616ec856043ca28cf9df Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Wed, 2 Sep 2026 17:45:23 -0500
Subject: [PATCH 22/36] set the file as world read so the docker user can read
 it

---
 check-project-links/action.yml | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 46593dc..71aef70 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -40,8 +40,10 @@ runs:
           rm -f "$link_file"
         }
         trap cleanup EXIT
+        chmod 0644 "$link_file"
 
         python3 "$GITHUB_ACTION_PATH/_find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
+        chmod 0644 "$link_file"
 
         if [ ! -s "$link_file" ]; then
           echo "No trusted project links found under $PROJECT_DIRECTORY."

From 970e7ad01de5a98796ce4bb63a8f3180475be0f8 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:12:54 -0500
Subject: [PATCH 23/36] extract comment searching to method

---
 check-project-links/_find_project_links.py | 35 +++++++++++-----------
 1 file changed, 18 insertions(+), 17 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index 0d9c5ce..a4e7c89 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -91,6 +91,23 @@ def _walk_metadata(value, results: set[str], allowed: set[str]) -> None:
             results.add(value)
 
 
+def _find_commented_links(manifest_path: str, results: set[str], allowed: set[str]) -> None:
+    """Find URLs embedded in comments within a pyproject.toml file."""
+    try:
+        with open(manifest_path, "r", encoding="utf-8") as manifest_file:
+            for line in manifest_file:
+                line = line.strip()
+                prefix, comment = line.split("#", maxsplit=1) if "#" in line else (line, "")
+                if comment.strip():
+                    comment = comment.strip()
+                    if comment.startswith("https://"):
+                        host = urlsplit(comment).hostname  # handles extra at the end just fine
+                        if host and _is_allowed(host, allowed):
+                            results.add(comment)
+    except OSError:
+        print(f"Warning: Failed to read pyproject.toml at {manifest_path}", file=sys.stderr)
+
+
 def _main() -> int:
     args = _parse_args()
     project_directory = args.project_directory
@@ -124,23 +141,7 @@ def _main() -> int:
                 continue
 
             _walk_metadata(metadata, results, allowed)
-            # Also find any commented URLs in the pyproject.toml file
-            try:
-                with open(manifest_path, "r", encoding="utf-8") as manifest_file:
-                    for line in manifest_file:
-                        line = line.strip()
-                        prefix, comment = line.split("#", maxsplit=1) if "#" in line else (line, "")
-                        if comment.strip():
-                            comment = comment.strip()
-                            if comment.startswith("https://"):
-                                host = urlsplit(
-                                    comment
-                                ).hostname  # handles extra at the end just fine
-                                if host and _is_allowed(host, allowed):
-                                    results.add(comment)
-            except OSError:
-                print(f"Warning: Failed to read pyproject.toml at {manifest_path}", file=sys.stderr)
-                continue
+            _find_commented_links(manifest_path, results, allowed)
 
     output_directory = os.path.dirname(safe_output_path)
     if output_directory:

From 97dbe0114e0a2334a5d58084932cbbead38c3f9b Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:13:30 -0500
Subject: [PATCH 24/36] remove duplicate check

---
 check-project-links/action.yml | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 71aef70..16a2ca8 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -53,14 +53,6 @@ runs:
         echo "Found $(wc -l < "$link_file") unique trusted links:"
         cat "$link_file"
 
-        if [ ! -s "$link_file" ]; then
-          echo "No trusted project links found."
-          exit 0
-        fi
-
-        echo "Found $(wc -l < "$link_file") unique trusted links:"
-        cat "$link_file"
-
         unprivileged_user=100:100 # this matches the default, but we want to be explicit about it
         docker run -u "${unprivileged_user}" --rm \
           -v "$link_file:/tmp/project_links.txt:ro" \

From 9ecc8aa5bb74b50836a8a1947291b6e3bd462eab Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:14:49 -0500
Subject: [PATCH 25/36] use RUNNER_TEMP

---
 check-project-links/action.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 16a2ca8..3768b4d 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -35,7 +35,7 @@ runs:
           exit 1
         fi
 
-        link_file="$(mktemp)"
+        link_file="$(mktemp -p \"$RUNNER_TEMP\")"
         cleanup() {
           rm -f "$link_file"
         }

From ee8f981b79c5f23fbbb517f3eae78e19b2631a6b Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:21:33 -0500
Subject: [PATCH 26/36] use RUNNER_TEMP and test not changing mode

---
 check-project-links/_find_project_links.py | 6 +++++-
 check-project-links/action.yml             | 4 +---
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index a4e7c89..fa2b429 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -114,7 +114,11 @@ def _main() -> int:
     safe_project_directory = os.path.realpath(project_directory, strict=True)
     safe_project_directory = _safe_under(os.getcwd(), safe_project_directory)
     allowed_domains = args.allowed_domains
-    safe_output_path = _safe_under("/tmp", os.path.realpath(args.output_path, strict=True))
+
+    runner_temp = os.environ.get("RUNNER_TEMP")
+    if not runner_temp:
+        raise RuntimeError("RUNNER_TEMP is required for output files.")
+    safe_output_path = _safe_under(runner_temp, os.path.realpath(args.output_path, strict=True))
 
     allowed: set[str] = set()
     for raw_domain in allowed_domains.split(","):
diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 3768b4d..cceae6c 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -35,15 +35,13 @@ runs:
           exit 1
         fi
 
-        link_file="$(mktemp -p \"$RUNNER_TEMP\")"
+        link_file="$(mktemp -p "$RUNNER_TEMP")"
         cleanup() {
           rm -f "$link_file"
         }
         trap cleanup EXIT
-        chmod 0644 "$link_file"
 
         python3 "$GITHUB_ACTION_PATH/_find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
-        chmod 0644 "$link_file"
 
         if [ ! -s "$link_file" ]; then
           echo "No trusted project links found under $PROJECT_DIRECTORY."

From 2da41f96a6bf037791846d1d4c42cf961831c7a9 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:28:03 -0500
Subject: [PATCH 27/36] turns out chmod is required because mktemp makes it too
 restricted

---
 check-project-links/action.yml | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index cceae6c..5a78e4f 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -41,6 +41,8 @@ runs:
         }
         trap cleanup EXIT
 
+        chmod 644 "$link_file"
+
         python3 "$GITHUB_ACTION_PATH/_find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
 
         if [ ! -s "$link_file" ]; then

From d69e2324f44dfa553ed3c22efcb6e7acf30611ac Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:34:51 -0500
Subject: [PATCH 28/36] don't require output to be in temp, just resolve it

---
 check-project-links/_find_project_links.py | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index fa2b429..177240f 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -114,11 +114,7 @@ def _main() -> int:
     safe_project_directory = os.path.realpath(project_directory, strict=True)
     safe_project_directory = _safe_under(os.getcwd(), safe_project_directory)
     allowed_domains = args.allowed_domains
-
-    runner_temp = os.environ.get("RUNNER_TEMP")
-    if not runner_temp:
-        raise RuntimeError("RUNNER_TEMP is required for output files.")
-    safe_output_path = _safe_under(runner_temp, os.path.realpath(args.output_path, strict=True))
+    safe_output_path = os.path.realpath(args.output_path, strict=True)
 
     allowed: set[str] = set()
     for raw_domain in allowed_domains.split(","):

From a24c0d843b627a5e147b67d09226dbae85f5688e Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:42:15 -0500
Subject: [PATCH 29/36] add type annotation

---
 check-project-links/_find_project_links.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index 177240f..b9085cd 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -4,6 +4,7 @@
 import ipaddress
 import os
 import sys
+import typing
 from urllib.parse import urlsplit
 
 try:
@@ -76,7 +77,9 @@ def _safe_under(base_dir: str, candidate_path: str) -> str:
     return safe_candidate
 
 
-def _walk_metadata(value, results: set[str], allowed: set[str]) -> None:
+def _walk_metadata(
+    value: typing.Union[dict, list, str], results: set[str], allowed: set[str]
+) -> None:
     if isinstance(value, dict):
         for item in value.values():
             _walk_metadata(item, results, allowed)

From e7461a426671c6613c4bdd48423c95870331c3c2 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:44:12 -0500
Subject: [PATCH 30/36] just say any

---
 check-project-links/_find_project_links.py | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index b9085cd..cd6584e 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -77,9 +77,7 @@ def _safe_under(base_dir: str, candidate_path: str) -> str:
     return safe_candidate
 
 
-def _walk_metadata(
-    value: typing.Union[dict, list, str], results: set[str], allowed: set[str]
-) -> None:
+def _walk_metadata(value: typing.Any, results: set[str], allowed: set[str]) -> None:
     if isinstance(value, dict):
         for item in value.values():
             _walk_metadata(item, results, allowed)

From c77494fd15aea2c27175a038c6f572c5a0ab9c75 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 10:50:50 -0500
Subject: [PATCH 31/36] let mypy pass on 3.10

---
 check-project-links/_find_project_links.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index cd6584e..29d3244 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -8,7 +8,7 @@
 from urllib.parse import urlsplit
 
 try:
-    import tomllib
+    import tomllib  # type: ignore[import-not-found]
 except ModuleNotFoundError:  # pragma: no cover
     raise Exception("tomllib is not available. Please use Python 3.11 or later.")
 

From ced8a580d68c10ac2c7fce76b1678eab3fc3185d Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 11:31:43 -0500
Subject: [PATCH 32/36] switch to just regex parsing

---
 check-project-links/_find_project_links.py | 60 +++++++---------------
 1 file changed, 19 insertions(+), 41 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index 29d3244..8b4e4a1 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -3,15 +3,10 @@
 import argparse
 import ipaddress
 import os
+import re
 import sys
-import typing
 from urllib.parse import urlsplit
 
-try:
-    import tomllib  # type: ignore[import-not-found]
-except ModuleNotFoundError:  # pragma: no cover
-    raise Exception("tomllib is not available. Please use Python 3.11 or later.")
-
 
 def _parse_args() -> argparse.Namespace:
     parser = argparse.ArgumentParser(
@@ -77,34 +72,28 @@ def _safe_under(base_dir: str, candidate_path: str) -> str:
     return safe_candidate
 
 
-def _walk_metadata(value: typing.Any, results: set[str], allowed: set[str]) -> None:
-    if isinstance(value, dict):
-        for item in value.values():
-            _walk_metadata(item, results, allowed)
-    elif isinstance(value, list):
-        for item in value:
-            _walk_metadata(item, results, allowed)
-    elif isinstance(value, str):
-        if not value.startswith("https://"):
-            return
-        host = urlsplit(value).hostname
-        if host and _is_allowed(host, allowed):
-            results.add(value)
+_URL_RE = re.compile(r'(https://.+?)(?:"|\s)')
+
+
+def _extract_links_from_line(line: str) -> list[str]:
+    """Extract URLs from a single line, stopping before a quote or whitespace."""
+    matches: list[str] = []
+    for match in re.finditer(_URL_RE, line):
+        url = match.group(1)
+        if url:
+            matches.append(url)
+    return matches
 
 
-def _find_commented_links(manifest_path: str, results: set[str], allowed: set[str]) -> None:
-    """Find URLs embedded in comments within a pyproject.toml file."""
+def _find_project_links(manifest_path: str, results: set[str], allowed: set[str]) -> None:
+    """Find URLs within a pyproject.toml file without parsing TOML."""
     try:
         with open(manifest_path, "r", encoding="utf-8") as manifest_file:
             for line in manifest_file:
-                line = line.strip()
-                prefix, comment = line.split("#", maxsplit=1) if "#" in line else (line, "")
-                if comment.strip():
-                    comment = comment.strip()
-                    if comment.startswith("https://"):
-                        host = urlsplit(comment).hostname  # handles extra at the end just fine
-                        if host and _is_allowed(host, allowed):
-                            results.add(comment)
+                for url in _extract_links_from_line(line):
+                    host = urlsplit(url).hostname
+                    if host and _is_allowed(host, allowed):
+                        results.add(url)
     except OSError:
         print(f"Warning: Failed to read pyproject.toml at {manifest_path}", file=sys.stderr)
 
@@ -131,18 +120,7 @@ def _main() -> int:
                 continue
 
             manifest_path = _safe_under(safe_project_directory, os.path.join(root, file_name))
-            try:
-                with open(manifest_path, "rb") as manifest_file:
-                    metadata = tomllib.load(manifest_file)
-            except (OSError, tomllib.TOMLDecodeError):
-                print(
-                    f"Warning: Failed to read or parse pyproject.toml at {manifest_path}",
-                    file=sys.stderr,
-                )
-                continue
-
-            _walk_metadata(metadata, results, allowed)
-            _find_commented_links(manifest_path, results, allowed)
+            _find_project_links(manifest_path, results, allowed)
 
     output_directory = os.path.dirname(safe_output_path)
     if output_directory:

From 99c7d0ba5b85fdd7a419dd9a124d0901d206b35a Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 11:39:27 -0500
Subject: [PATCH 33/36] add comment

---
 check-project-links/action.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index 5a78e4f..c342a8f 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -41,7 +41,7 @@ runs:
         }
         trap cleanup EXIT
 
-        chmod 644 "$link_file"
+        chmod 644 "$link_file" # make the file world-readable so that the unprivileged user in the docker container can read it
 
         python3 "$GITHUB_ACTION_PATH/_find_project_links.py" "$PROJECT_DIRECTORY" "$ALLOWED_DOMAINS" "$link_file"
 

From e410a1c73b022f8589e0476351e061b199d83ac2 Mon Sep 17 00:00:00 2001
From: mshafer-NI <23644905+mshafer-NI@users.noreply.github.com>
Date: Thu, 3 Sep 2026 12:41:25 -0500
Subject: [PATCH 34/36] Apply batched suggestions from code review

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
---
 check-project-links/README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/check-project-links/README.md b/check-project-links/README.md
index bcb4a67..6a03ff7 100644
--- a/check-project-links/README.md
+++ b/check-project-links/README.md
@@ -36,7 +36,7 @@ steps:
   - uses: actions/checkout@v0
 
   - name: Check project links
-    uses: ni/python-actions/check-project-links@v1
+    uses: ni/python-actions/check-project-links@v0
     with:
       project-directory: .
       docker-image: curlimages/curl:8.22.0
@@ -44,8 +44,8 @@ steps:
 
 ## Behavior
 
-- Uses Python directory walk and [`tomllib`](https://docs.python.org/3/library/tomllib.html) + string search to identify all links in pyproject.toml files under the provided path.
-- Extracts strings that are parsable URLs, and URLs from comments
+- Uses Python directory walk + regex search to identify `https://` links in `pyproject.toml` files under the provided path.
+- Extracts `https://` substrings (including from comments) and filters them by the configured allowed domains before validation.
 - Drops any URL whose hostname is `localhost`, a local loopback address, or any literal IP address before validation.
 - Deduplicates the list of URLs and writes them to a temporary file.
 - Validates each URL with the configured Docker image.

From 4b805eb92e902587b1c74c2698d6f9d406ee1bf4 Mon Sep 17 00:00:00 2001
From: mshafer-NI <23644905+mshafer-NI@users.noreply.github.com>
Date: Thu, 3 Sep 2026 12:41:59 -0500
Subject: [PATCH 35/36] Simplify exit handling for project link checks

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
---
 check-project-links/action.yml | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/check-project-links/action.yml b/check-project-links/action.yml
index c342a8f..2055a9c 100644
--- a/check-project-links/action.yml
+++ b/check-project-links/action.yml
@@ -78,9 +78,6 @@ runs:
             exit "$failed"
           '
 
-        status=$?
-        if [ "$status" -ne 0 ]; then
-          exit 1
-        fi
+        # `docker run` failures already cause this step to fail due to `set -e`.
 
         echo "Success: No checked project links returned 4xx responses."

From 92c4c6ea66aa5b165b1c7be1a98b97731f6a92c8 Mon Sep 17 00:00:00 2001
From: Matthew Shafer <matthew.shafer@emerson.com>
Date: Thu, 3 Sep 2026 12:45:49 -0500
Subject: [PATCH 36/36] tone down the sanitizing

---
 check-project-links/_find_project_links.py | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/check-project-links/_find_project_links.py b/check-project-links/_find_project_links.py
index 8b4e4a1..5b99398 100644
--- a/check-project-links/_find_project_links.py
+++ b/check-project-links/_find_project_links.py
@@ -100,11 +100,9 @@ def _find_project_links(manifest_path: str, results: set[str], allowed: set[str]
 
 def _main() -> int:
     args = _parse_args()
-    project_directory = args.project_directory
-    safe_project_directory = os.path.realpath(project_directory, strict=True)
-    safe_project_directory = _safe_under(os.getcwd(), safe_project_directory)
+    safe_project_directory = _safe_under(os.getcwd(), args.project_directory)
     allowed_domains = args.allowed_domains
-    safe_output_path = os.path.realpath(args.output_path, strict=True)
+    output_path = args.output_path
 
     allowed: set[str] = set()
     for raw_domain in allowed_domains.split(","):
@@ -122,11 +120,11 @@ def _main() -> int:
             manifest_path = _safe_under(safe_project_directory, os.path.join(root, file_name))
             _find_project_links(manifest_path, results, allowed)
 
-    output_directory = os.path.dirname(safe_output_path)
+    output_directory = os.path.dirname(output_path)
     if output_directory:
         os.makedirs(output_directory, exist_ok=True)
 
-    with open(safe_output_path, "w", encoding="utf-8") as output_file:
+    with open(output_path, "w", encoding="utf-8") as output_file:
         for url in sorted(results):
             output_file.write(f"{url}\n")