Files
review-bot/.gitea/actions/review/action.yml
T
Rodin c349986187
PR Ready Gate / clear-labels (pull_request) Successful in 2s
CI / test (pull_request) Successful in 25s
CI / review (anthropic--claude-4.6-sonnet, sonnet, SONNET_REVIEW_TOKEN) (pull_request) Successful in 41s
CI / review (gpt-5, gpt, GPT_REVIEW_TOKEN) (pull_request) Successful in 2m16s
CI / review (gpt-5, security, ., rodin/security-patterns, SECURITY_REVIEW.md, SECURITY_REVIEW_TOKEN) (pull_request) Successful in 2m49s
fix(#123): add RFC6598 CGN check to Python SSRF validation in action.yml
Python's ipaddress module does NOT classify 100.64.0.0/10 (RFC6598
carrier-grade NAT) as private/loopback/link_local/multicast/reserved.
This means a SERVER_URL resolving to a CGN address would bypass the
Python SSRF check and reach curl with ACTION_TOKEN.

Add an explicit network membership check for 100.64.0.0/10 to both
Python validation blocks in action.yml:
  - _ssrf_check.py (VCS URL pre-flight check)
  - _ssrf_check_install.py (binary download URL check)

The Go-level IsBlockedIP already covers this range correctly (ipcheck.go);
this fix closes the gap in the action.yml Python layer.

Also update comments to mention RFC6598 explicitly.
2026-05-14 13:41:17 +00:00

507 lines
23 KiB
YAML

# This composite action supports both Gitea Actions and GitHub Actions runners.
# It detects the VCS host type by checking whether github.api_url is set
# (present on GitHub.com and GHES runners, absent on Gitea runners) and uses
# the appropriate releases API for version resolution and binary download
# (REST API on GitHub, direct URLs on Gitea).
#
# Security notes:
# - On GitHub/GHES (VCS_TYPE=github), inputs.vcs-url is IGNORED to prevent
# token exfiltration. API calls use github.api_url; downloads use
# github.server_url. Tokens are never sent to user-supplied URLs.
# - On Gitea (VCS_TYPE=gitea), inputs.vcs-url is validated (https scheme,
# no whitespace/newlines, and DNS resolution to a public IP) before use.
# Python3 resolves the hostname and rejects RFC1918, RFC6598 (carrier-grade
# NAT), loopback, link-local, and other reserved addresses to prevent SSRF attacks.
# The installed review-bot binary additionally uses a safe HTTP transport
# (DialContext-level IP check) for all Gitea API calls at runtime.
# The binary also exposes a `validate-url` subcommand for use in any future
# shell steps that need to validate a URL before passing it to curl.
# - action-repo is validated against owner/repo pattern.
# - Tokens are passed via masked environment variables, not step outputs.
#
# Requirements: python3, sha256sum, curl (all present on ubuntu-* runners).
name: 'AI Code Review'
description: 'Run AI-powered code review on a pull request using review-bot'
inputs:
vcs-url:
description: 'VCS server URL (only used on Gitea runners; ignored on GitHub/GHES). Defaults to server_url.'
required: false
default: ''
repo:
description: 'Repository to review (owner/name, defaults to current)'
required: false
default: ''
action-repo:
description: 'Repository hosting review-bot releases (owner/name). Defaults to github.action_repository or rodin/review-bot.'
required: false
default: ''
action-repo-token:
description: 'Token for downloading release assets from action-repo (defaults to github.token on GitHub, reviewer-token on Gitea). Required for private repos.'
required: false
default: ''
pr-number:
description: 'Pull request number (defaults to current PR)'
required: false
default: ''
reviewer-token:
description: 'Token for posting the review'
required: true
reviewer-name:
description: 'Display name for the reviewer'
required: false
default: ''
llm-base-url:
description: 'OpenAI-compatible LLM API base URL (not required for aicore provider)'
required: false
default: ''
llm-api-key:
description: 'LLM API key (not required for aicore provider)'
required: false
default: ''
llm-model:
description: 'LLM model name'
required: true
llm-provider:
description: 'LLM API provider: openai, anthropic, or aicore (default openai)'
required: false
default: 'openai'
aicore-client-id:
description: 'SAP AI Core client ID (required for aicore provider)'
required: false
default: ''
aicore-client-secret:
description: 'SAP AI Core client secret (required for aicore provider)'
required: false
default: ''
aicore-auth-url:
description: 'SAP AI Core authentication URL (required for aicore provider)'
required: false
default: ''
aicore-api-url:
description: 'SAP AI Core API URL (required for aicore provider)'
required: false
default: ''
aicore-resource-group:
description: 'SAP AI Core resource group (default: default)'
required: false
default: 'default'
conventions-file:
description: 'Path to conventions file in the repo (e.g. CLAUDE.md)'
required: false
default: ''
patterns-repo:
description: 'Comma-separated repos with language patterns (e.g. rodin/elixir-patterns,rodin/phoenix-conventions)'
required: false
default: ''
patterns-files:
description: 'Comma-separated file paths or directories to fetch from patterns repos'
required: false
default: 'README.md'
temperature:
description: 'LLM temperature (0 = server default)'
required: false
default: '0'
timeout:
description: 'LLM request timeout in seconds (default 300)'
required: false
default: '300'
version:
description: 'review-bot version to install (e.g. v0.1.0, defaults to latest)'
required: false
default: 'latest'
dry-run:
description: 'Print review to stdout instead of posting'
required: false
default: 'false'
update-existing:
description: 'Delete previous review from same bot after posting new one. Accepts: true/1/yes or false/0/no (default true)'
required: false
default: 'true'
system-prompt-file:
description: 'Local file with additional system prompt instructions (e.g. security review focus)'
required: false
default: ''
persona:
description: 'Built-in persona name (security, architect, docs)'
required: false
default: ''
persona-file:
description: 'Path to custom persona JSON file'
required: false
default: ''
runs:
using: 'composite'
steps:
- name: Determine version
id: version
shell: bash
run: |
set -euo pipefail
# --- Input Validation ---
# Determine the repo hosting review-bot releases (not the repo being reviewed)
ACTION_REPO="${{ inputs.action-repo }}"
if [ -z "$ACTION_REPO" ]; then
# github.action_repository is the repo containing the running action
ACTION_REPO="${{ github.action_repository }}"
fi
if [ -z "$ACTION_REPO" ]; then
# Final fallback for Gitea (which may not set action_repository)
ACTION_REPO="rodin/review-bot"
echo "::notice::action-repo not specified and github.action_repository is empty; falling back to rodin/review-bot"
fi
# Validate ACTION_REPO matches owner/repo pattern (prevent path traversal)
if ! printf '%s' "$ACTION_REPO" | grep -qE '^[a-zA-Z0-9._-]+/[a-zA-Z0-9._-]+$'; then
echo "Error: action-repo '${ACTION_REPO}' does not match expected owner/repo format" >&2
exit 1
fi
# Detect VCS host type using github.api_url context.
# github.api_url is set on GitHub.com (https://api.github.com) and GHES
# (https://<host>/api/v3). It is empty/unset on Gitea Actions runners.
GITHUB_API_URL="${{ github.api_url }}"
if [ -n "$GITHUB_API_URL" ]; then
VCS_TYPE="github"
else
VCS_TYPE="gitea"
fi
# Determine SERVER_URL based on VCS type.
# SECURITY: On GitHub/GHES, ALWAYS use github.server_url — never trust
# inputs.vcs-url to prevent token exfiltration to attacker-controlled hosts.
if [ "$VCS_TYPE" = "github" ]; then
SERVER_URL="${{ github.server_url }}"
if [ -n "${{ inputs.vcs-url }}" ]; then
echo "::warning::inputs.vcs-url is ignored on GitHub/GHES runners (VCS_TYPE=github). Using github.server_url instead."
fi
else
SERVER_URL="${{ inputs.vcs-url || github.server_url }}"
fi
# Strip trailing slash if present
SERVER_URL="${SERVER_URL%/}"
# Validate SERVER_URL for Gitea path: must be https, no whitespace/newlines.
# The [^[:space:]] class already rejects newlines, so no separate newline check needed.
if [ "$VCS_TYPE" = "gitea" ]; then
if ! printf '%s' "$SERVER_URL" | grep -qE '^https://[^[:space:]]+$'; then
echo "Error: SERVER_URL '${SERVER_URL}' must be an https:// URL with no whitespace" >&2
exit 1
fi
# Additional IP-level SSRF defense: resolve the hostname and reject
# requests to RFC1918, RFC6598 (carrier-grade NAT), loopback, link-local,
# and other reserved addresses.
# python3 is required on ubuntu-* runners (see requirements comment above).
# Use printf to write the script to a temp file so the python lines are valid
# YAML (each indented line becomes a printf argument — no unindented code).
# SERVER_URL is passed via CHECK_URL env var, never interpolated into python code.
printf '%s\n' \
'import socket,ipaddress,sys,os' \
'from urllib.parse import urlparse' \
'u=os.environ["CHECK_URL"]; parsed=urlparse(u)' \
'if parsed.username or parsed.password:' \
' print("Error: URL contains user-info — not allowed",file=sys.stderr); sys.exit(2)' \
'h=parsed.hostname' \
'(print("Error: no hostname",file=sys.stderr) or sys.exit(2)) if not h else None' \
'try: rs=socket.getaddrinfo(h,None)' \
'except socket.gaierror as e: print(f"DNS error: {e}",file=sys.stderr); sys.exit(1)' \
'if not rs: print("Error: no addresses",file=sys.stderr); sys.exit(1)' \
'for _,_,_,_,(a,*_) in rs:' \
' ip=ipaddress.ip_address(a)' \
' if isinstance(ip,ipaddress.IPv6Address) and ip.ipv4_mapped: ip=ip.ipv4_mapped' \
' cgn=ipaddress.ip_network("100.64.0.0/10")' \
' if ip.is_private or ip.is_loopback or ip.is_link_local or ip.is_multicast or ip.is_reserved or ip in cgn:' \
' print(f"blocked: {a}",file=sys.stderr); sys.exit(1)' \
> /tmp/_ssrf_check.py
CHECK_URL="${SERVER_URL}" python3 /tmp/_ssrf_check.py || {
echo "Error: SERVER_URL '${SERVER_URL}' resolves to a private/reserved IP address" >&2
exit 1
}
fi
# Determine auth token for release API requests
ACTION_TOKEN="${{ inputs.action-repo-token }}"
if [ -z "$ACTION_TOKEN" ]; then
if [ "$VCS_TYPE" = "github" ]; then
ACTION_TOKEN="${{ github.token }}"
else
ACTION_TOKEN="${{ inputs.reviewer-token }}"
fi
fi
# Validate token contains no control characters (defense-in-depth against header injection)
if [ -n "$ACTION_TOKEN" ]; then
if printf '%s' "$ACTION_TOKEN" | LC_ALL=C grep -q '[^[:print:]]'; then
echo "Error: ACTION_TOKEN contains control characters" >&2
exit 1
fi
fi
if [ "${{ inputs.version }}" = "latest" ]; then
if [ "$VCS_TYPE" = "github" ]; then
# SECURITY: Use github.api_url which is a trusted platform-provided value.
# Never construct API URLs from user-supplied inputs on GitHub.
API_URL="${GITHUB_API_URL}/repos/${ACTION_REPO}/releases?per_page=1"
else
# Gitea API — SERVER_URL was validated above
API_URL="${SERVER_URL}/api/v1/repos/${ACTION_REPO}/releases?limit=1"
fi
# Fetch latest version with inline auth header (no intermediate variable)
if [ -n "$ACTION_TOKEN" ]; then
if [ "$VCS_TYPE" = "github" ]; then
VERSION=$(curl -sSf --connect-timeout 10 --max-time 30 \
-H "Authorization: Bearer ${ACTION_TOKEN}" "$API_URL" \
| python3 -c "import sys, json; releases = json.load(sys.stdin); print(releases[0]['tag_name'] if releases else '')")
else
VERSION=$(curl -sSf --connect-timeout 10 --max-time 30 \
-H "Authorization: token ${ACTION_TOKEN}" "$API_URL" \
| python3 -c "import sys, json; releases = json.load(sys.stdin); print(releases[0]['tag_name'] if releases else '')")
fi
else
VERSION=$(curl -sSf --connect-timeout 10 --max-time 30 "$API_URL" \
| python3 -c "import sys, json; releases = json.load(sys.stdin); print(releases[0]['tag_name'] if releases else '')")
fi
if [ -z "$VERSION" ]; then
echo "Failed to determine latest version from ${API_URL}" >&2
exit 1
fi
else
VERSION="${{ inputs.version }}"
fi
# Validate VERSION: no slashes or whitespace (prevent path traversal).
# [:space:] includes newlines and carriage returns in POSIX.
if printf '%s' "$VERSION" | grep -qE '[/[:space:]]'; then
echo "Error: VERSION '${VERSION}' contains invalid characters (newline, slash, or whitespace)" >&2
exit 1
fi
# Detect OS and architecture for platform-specific binary download
OS_RAW=$(uname -s | tr '[:upper:]' '[:lower:]')
case "$OS_RAW" in
linux) OS="linux" ;;
darwin) OS="darwin" ;;
*)
echo "Error: unsupported OS: $(uname -s)" >&2
exit 1
;;
esac
RAW_ARCH=$(uname -m)
case "$RAW_ARCH" in
x86_64) ARCH="amd64" ;;
aarch64 | arm64) ARCH="arm64" ;;
*)
echo "Error: unsupported architecture: $RAW_ARCH" >&2
exit 1
;;
esac
echo "version=${VERSION}" >> "$GITHUB_OUTPUT"
echo "os=${OS}" >> "$GITHUB_OUTPUT"
echo "arch=${ARCH}" >> "$GITHUB_OUTPUT"
echo "action_repo=${ACTION_REPO}" >> "$GITHUB_OUTPUT"
echo "server_url=${SERVER_URL}" >> "$GITHUB_OUTPUT"
echo "vcs_type=${VCS_TYPE}" >> "$GITHUB_OUTPUT"
# SECURITY: Pass token via masked environment variable instead of step output.
# Step outputs can leak in debug logs; GITHUB_ENV with masking is safer.
if [ -n "$ACTION_TOKEN" ]; then
echo "::add-mask::${ACTION_TOKEN}"
echo "ACTION_TOKEN=${ACTION_TOKEN}" >> "$GITHUB_ENV"
fi
- name: Cache review-bot binary
id: cache
uses: actions/cache@v4
with:
path: ${{ runner.temp }}/review-bot
key: review-bot-${{ steps.version.outputs.os }}-${{ steps.version.outputs.arch }}-${{ steps.version.outputs.version }}
- name: Install review-bot
if: steps.cache.outputs.cache-hit != 'true'
shell: bash
run: |
set -euo pipefail
SERVER_URL="${{ steps.version.outputs.server_url }}"
ACTION_REPO="${{ steps.version.outputs.action_repo }}"
VERSION="${{ steps.version.outputs.version }}"
VCS_TYPE="${{ steps.version.outputs.vcs_type }}"
OS="${{ steps.version.outputs.os }}"
ARCH="${{ steps.version.outputs.arch }}"
# Read token from masked environment variable (set in Determine version step)
# Falls back to empty if not set (public repos don't need auth)
ACTION_TOKEN="${ACTION_TOKEN:-}"
BINARY="review-bot-${OS}-${ARCH}"
# SECURITY: Re-validate SERVER_URL at the start of this step to mitigate DNS
# rebinding attacks. A DNS TTL expiry between "Determine version" and here
# could allow an attacker to change the resolved IP to a private/reserved
# address, causing curl to send ACTION_TOKEN to an internal host.
# Only needed on Gitea path (VCS_TYPE=gitea); GitHub/GHES uses platform-controlled URLs.
if [ "$VCS_TYPE" = "gitea" ]; then
printf '%s\n' \
'import socket,ipaddress,sys,os' \
'from urllib.parse import urlparse' \
'u=os.environ["CHECK_URL"]; parsed=urlparse(u)' \
'if parsed.username or parsed.password:' \
' print("Error: URL contains user-info — not allowed",file=sys.stderr); sys.exit(2)' \
'h=parsed.hostname' \
'(print("Error: no hostname",file=sys.stderr) or sys.exit(2)) if not h else None' \
'try: rs=socket.getaddrinfo(h,None)' \
'except socket.gaierror as e: print(f"DNS error: {e}",file=sys.stderr); sys.exit(1)' \
'if not rs: print("Error: no addresses",file=sys.stderr); sys.exit(1)' \
'for _,_,_,_,(a,*_) in rs:' \
' ip=ipaddress.ip_address(a)' \
' if isinstance(ip,ipaddress.IPv6Address) and ip.ipv4_mapped: ip=ip.ipv4_mapped' \
' cgn=ipaddress.ip_network("100.64.0.0/10")' \
' if ip.is_private or ip.is_loopback or ip.is_link_local or ip.is_multicast or ip.is_reserved or ip in cgn:' \
' print(f"blocked: {a}",file=sys.stderr); sys.exit(1)' \
> /tmp/_ssrf_check_install.py
CHECK_URL="${SERVER_URL}" python3 /tmp/_ssrf_check_install.py || {
echo "Error: SERVER_URL '${SERVER_URL}' resolves to a private/reserved IP address" >&2
exit 1
}
fi
if [ "$VCS_TYPE" = "github" ]; then
# GitHub/GHES: Use REST API for release asset downloads.
# Web release URLs ({server}/.../releases/download/{tag}/{asset}) redirect
# to S3 and don't reliably support Authorization headers for private repos.
# The REST API endpoint with Accept: application/octet-stream is required.
# GITHUB_API_URL: trusted platform value, same as detected in "Determine version" step.
GITHUB_API_URL="${{ github.api_url }}"
if [ -n "$ACTION_TOKEN" ]; then
RELEASE_JSON=$(curl -sSf --connect-timeout 10 --max-time 30 \
-H "Authorization: Bearer ${ACTION_TOKEN}" \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/tags/${VERSION}")
else
RELEASE_JSON=$(curl -sSf --connect-timeout 10 --max-time 30 \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/tags/${VERSION}")
fi
# Extract asset IDs for binary and checksums
BINARY_ASSET_ID=$(printf '%s' "$RELEASE_JSON" | python3 -c "import sys, json; assets = json.load(sys.stdin).get('assets', []); matches = [a['id'] for a in assets if a['name'] == '${BINARY}']; print(matches[0] if matches else '')")
if [ -z "$BINARY_ASSET_ID" ]; then
echo "Error: could not find asset '${BINARY}' in release ${VERSION}" >&2
exit 1
fi
CHECKSUMS_ASSET_ID=$(printf '%s' "$RELEASE_JSON" | python3 -c "import sys, json; assets = json.load(sys.stdin).get('assets', []); matches = [a['id'] for a in assets if a['name'] == 'checksums.txt']; print(matches[0] if matches else '')")
if [ -z "$CHECKSUMS_ASSET_ID" ]; then
echo "Error: could not find asset 'checksums.txt' in release ${VERSION}" >&2
exit 1
fi
# Download assets via REST API with Accept: application/octet-stream
if [ -n "$ACTION_TOKEN" ]; then
curl -sSfL --connect-timeout 10 --max-time 120 \
-H "Authorization: Bearer ${ACTION_TOKEN}" \
-H "Accept: application/octet-stream" \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/assets/${BINARY_ASSET_ID}" \
-o "${{ runner.temp }}/review-bot"
curl -sSfL --connect-timeout 10 --max-time 30 \
-H "Authorization: Bearer ${ACTION_TOKEN}" \
-H "Accept: application/octet-stream" \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/assets/${CHECKSUMS_ASSET_ID}" \
-o "${{ runner.temp }}/checksums.txt"
else
curl -sSfL --connect-timeout 10 --max-time 120 \
-H "Accept: application/octet-stream" \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/assets/${BINARY_ASSET_ID}" \
-o "${{ runner.temp }}/review-bot"
curl -sSfL --connect-timeout 10 --max-time 30 \
-H "Accept: application/octet-stream" \
"${GITHUB_API_URL}/repos/${ACTION_REPO}/releases/assets/${CHECKSUMS_ASSET_ID}" \
-o "${{ runner.temp }}/checksums.txt"
fi
else
# Gitea: Direct download via web release URLs (Gitea serves assets
# directly without redirects — no -L needed).
# SECURITY: Omitting -L prevents forwarding Authorization header to
# unexpected hosts if Gitea ever introduces CDN redirects.
DOWNLOAD_URL="${SERVER_URL}/${ACTION_REPO}/releases/download/${VERSION}"
if [ -n "$ACTION_TOKEN" ]; then
curl -sSf --connect-timeout 10 --max-time 120 \
-H "Authorization: token ${ACTION_TOKEN}" \
"${DOWNLOAD_URL}/${BINARY}" -o "${{ runner.temp }}/review-bot"
curl -sSf --connect-timeout 10 --max-time 30 \
-H "Authorization: token ${ACTION_TOKEN}" \
"${DOWNLOAD_URL}/checksums.txt" -o "${{ runner.temp }}/checksums.txt"
else
curl -sSf --connect-timeout 10 --max-time 120 \
"${DOWNLOAD_URL}/${BINARY}" -o "${{ runner.temp }}/review-bot"
curl -sSf --connect-timeout 10 --max-time 30 \
"${DOWNLOAD_URL}/checksums.txt" -o "${{ runner.temp }}/checksums.txt"
fi
fi
# Verify SHA-256 checksum
# NOTE: This verifies integrity (download wasn't corrupted) but not
# authenticity — both binary and checksums come from the same server.
# For stronger guarantees, consider GPG signature verification.
cd "${{ runner.temp }}"
EXPECTED=$(grep -E "^[0-9a-f]+[[:space:]]+\*?${BINARY}$" checksums.txt | awk '{print $1}')
# sha256sum (GNU) is not available on macOS; use shasum -a 256 on darwin.
if [ "${OS}" = "darwin" ]; then
ACTUAL=$(shasum -a 256 review-bot | awk '{print $1}')
else
ACTUAL=$(sha256sum review-bot | awk '{print $1}')
fi
if [ -z "$EXPECTED" ]; then
echo "Error: no checksum found for ${BINARY}" >&2
exit 1
fi
if [ "$EXPECTED" != "$ACTUAL" ]; then
echo "Error: checksum mismatch!" >&2
echo " Expected: $EXPECTED" >&2
echo " Actual: $ACTUAL" >&2
exit 1
fi
chmod +x "${{ runner.temp }}/review-bot"
echo "Installed review-bot-${OS}-${ARCH} ${VERSION} (checksum verified)"
- name: Run review
shell: bash
env:
VCS_URL: ${{ steps.version.outputs.server_url }}
GITEA_REPO: ${{ inputs.repo || github.repository }}
PR_NUMBER: ${{ inputs.pr-number || github.event.pull_request.number }}
REVIEWER_TOKEN: ${{ inputs.reviewer-token }}
REVIEWER_NAME: ${{ inputs.reviewer-name }}
LLM_BASE_URL: ${{ inputs.llm-base-url }}
LLM_API_KEY: ${{ inputs.llm-api-key }}
LLM_MODEL: ${{ inputs.llm-model }}
CONVENTIONS_FILE: ${{ inputs.conventions-file }}
PATTERNS_REPO: ${{ inputs.patterns-repo }}
PATTERNS_FILES: ${{ inputs.patterns-files }}
LLM_TEMPERATURE: ${{ inputs.temperature }}
LLM_TIMEOUT: ${{ inputs.timeout }}
LLM_PROVIDER: ${{ inputs.llm-provider }}
UPDATE_EXISTING: ${{ inputs.update-existing }}
SYSTEM_PROMPT_FILE: ${{ inputs.system-prompt-file }}
PERSONA: ${{ inputs.persona }}
PERSONA_FILE: ${{ inputs.persona-file }}
AICORE_CLIENT_ID: ${{ inputs.aicore-client-id }}
AICORE_CLIENT_SECRET: ${{ inputs.aicore-client-secret }}
AICORE_AUTH_URL: ${{ inputs.aicore-auth-url }}
AICORE_API_URL: ${{ inputs.aicore-api-url }}
AICORE_RESOURCE_GROUP: ${{ inputs.aicore-resource-group }}
run: |
ARGS=""
if [ "${{ inputs.dry-run }}" = "true" ]; then
ARGS="--dry-run"
fi
${{ runner.temp }}/review-bot $ARGS