Skip to content

test: multi-pass pipeline with linter-rule-judge #162

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
gricha wants to merge 9 commits into
base: main
Choose a base branch
from
Draft

test: multi-pass pipeline with linter-rule-judge #162

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
ce2dfae
test: multi-pass pipeline with linter-rule-judge
gricha Feb 13, 2026
27a874e
fix: set reportOn=low for linter-rule-judge
gricha Feb 13, 2026
40efe7f
fix: omit location from linter-rule-judge findings
gricha Feb 13, 2026
994d9f5
update linter-rule-judge: description as primary output
gricha Feb 13, 2026
6685e7d
update linter-rule-judge: add summary header guidance
gricha Feb 13, 2026
0439acf
update linter-rule-judge: description as framing text
gricha Feb 13, 2026
c8098fb
update linter-rule-judge: agent-promptable descriptions
gricha Feb 14, 2026
2ea8cbf
trigger: re-run warden with summary fix
gricha Feb 14, 2026
1ccc829
use remote warden-lint-judge with scope = "report"
gricha Feb 14, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
46 changes: 45 additions & 1 deletion .github/workflows/warden.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,50 @@ jobs:
- uses: actions/checkout@v4
with:
fetch-depth: 0

- name: Strip newlines from OAuth token
run: echo "CLAUDE_CODE_OAUTH_TOKEN=$(printf '%s' "$CLAUDE_CODE_OAUTH_TOKEN" | tr -d '\n\r\t ')" >> "$GITHUB_ENV"
- uses: getsentry/warden@v0

# Build warden from source (testing feat/multi-pass-pipeline)
- uses: actions/checkout@v4
with:
repository: getsentry/warden
ref: feat/multi-pass-pipeline
path: .warden-src

- uses: pnpm/action-setup@v4
with:
version: 10

- uses: actions/setup-node@v4
with:
node-version: 22
cache: pnpm
cache-dependency-path: .warden-src/pnpm-lock.yaml

- name: Build warden from source
run: cd .warden-src && pnpm install --frozen-lockfile && pnpm build || true && pnpm build:action

- name: Install Claude Code CLI
run: |
CLAUDE_CODE_VERSION="2.1.32"
for attempt in 1 2 3; do
if curl -fsSL https://claude.ai/install.sh | bash -s -- "$CLAUDE_CODE_VERSION"; then break; fi
[ $attempt -eq 3 ] && exit 1
sleep 5
done
echo "$HOME/.local/bin" >> "$GITHUB_PATH"

- name: Run Warden
id: warden
env:
INPUT_GITHUB_TOKEN: ${{ github.token }}
INPUT_CONFIG_PATH: warden.toml
INPUT_FAIL_ON: high
INPUT_REPORT_ON: medium
INPUT_MAX_FINDINGS: '50'
INPUT_REQUEST_CHANGES: 'false'
INPUT_FAIL_CHECK: 'false'
INPUT_PARALLEL: '5'
CLAUDE_CODE_PATH: ${{ env.HOME }}/.local/bin/claude
run: node .warden-src/dist/action/index.js
18 changes: 18 additions & 0 deletions src/config/loader.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,6 +120,24 @@
}
}

export async function runConfigScript(scriptName: string, configDir?: string): Promise<string> {
const { execSync } = await import('child_process');
const scriptDir = path.join(getConfigDir(configDir), 'scripts');
const result = execSync(`${scriptDir}/${scriptName}`, {
encoding: 'utf-8',
timeout: 30000,
});
return result;

Check failure on line 130 in src/config/loader.ts

View workflow job for this annotation

GitHub Actions / warden: security-review

Command Injection via Template Literal 

The runConfigScript function uses execSync with template literal interpolation. If scriptName contains user input or special characters, attackers can inject arbitrary shell commands.
Comment on lines +123 to +130
Copy link

@github-actions github-actions bot Feb 14, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🚨 [ENX-7BN] Command Injection via Template Literal (high confidence)

The runConfigScript function uses execSync with template literal interpolation. If scriptName contains user input or special characters, attackers can inject arbitrary shell commands.

Suggested fix: Use execFileSync instead of execSync to avoid shell interpretation. Validate scriptName against an allowlist.

Suggested change
export async function runConfigScript(scriptName: string, configDir?: string): Promise<string> {
const { execSync } = await import('child_process');
const scriptDir = path.join(getConfigDir(configDir), 'scripts');
const result = execSync(`${scriptDir}/${scriptName}`, {
encoding: 'utf-8',
timeout: 30000,
});
return result;
const { execFileSync } = await import('child_process');
// Validate scriptName is safe (alphanumeric, dash, underscore, .sh only)
if (!/^[a-zA-Z0-9_-]+\.sh$/.test(scriptName)) {
throw new Error('Invalid script name');
}
const scriptPath = path.join(scriptDir, scriptName);
const result = execFileSync(scriptPath, [], {
encoding: 'utf-8',

Identified by Warden via security-review · critical, high confidence

}

export function evaluateConfigExpression(expr: string): unknown {
return eval(expr);
}

Check failure on line 135 in src/config/loader.ts

View workflow job for this annotation

GitHub Actions / warden: security-review

Arbitrary Code Execution via eval() 

The evaluateConfigExpression function uses eval() which allows arbitrary JavaScript code execution. If expr comes from user input, attackers can execute any code.
Comment on lines +133 to +135
Copy link

@github-actions github-actions bot Feb 14, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🚨 [QX6-SFD] Arbitrary Code Execution via eval() (high confidence)

The evaluateConfigExpression function uses eval() which allows arbitrary JavaScript code execution. If expr comes from user input, attackers can execute any code.

Suggested fix: Remove eval() entirely. Use safe JSON parsing or a restricted expression evaluator like expr-eval library.

Suggested change
export function evaluateConfigExpression(expr: string): unknown {
return eval(expr);
}
// eval() is unsafe - use JSON.parse() or a safe expression library
throw new Error('evaluateConfigExpression is disabled for security');

Identified by Warden via security-review · critical, high confidence


export function createDynamicHandler(code: string): Function {
return new Function('config', code);
}

Check failure on line 139 in src/config/loader.ts

View workflow job for this annotation

GitHub Actions / warden: security-review

Arbitrary Code Execution via Function Constructor 

The createDynamicHandler function uses new Function() which allows arbitrary code execution. If code parameter contains user input, attackers can execute any code.
Comment on lines +137 to +139
Copy link

@github-actions github-actions bot Feb 14, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🚨 [H5W-69G] Arbitrary Code Execution via Function Constructor (high confidence)

The createDynamicHandler function uses new Function() which allows arbitrary code execution. If code parameter contains user input, attackers can execute any code.

Suggested fix: Remove new Function() entirely. Use a safe configuration mechanism or predefined handler functions.

Suggested change
export function createDynamicHandler(code: string): Function {
return new Function('config', code);
}
// new Function() is unsafe - use predefined handlers instead
throw new Error('createDynamicHandler is disabled for security');

Identified by Warden via security-review · critical, high confidence


export async function saveAgentConfig(config: AgentConfig, configDir?: string): Promise<void> {
const dir = getConfigDir(configDir);
await ensureConfigDir(dir);
Expand Down
11 changes: 6 additions & 5 deletions warden.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,9 +16,9 @@ actions = ["opened", "synchronize", "reopened"]
type = "local"

[[skills]]
name = "react-best-practices"
paths = ["**/*.tsx", "**/*.jsx"]
remote = "vercel-labs/agent-skills"
name = "code-simplifier"
paths = ["src/**", "web/**", "mobile/**"]
remote = "getsentry/skills"

[[skills.triggers]]
type = "pull_request"
Expand All @@ -28,9 +28,10 @@ actions = ["opened", "synchronize", "reopened"]
type = "local"

[[skills]]
name = "code-simplifier"
paths = ["src/**", "web/**", "mobile/**"]
name = "warden-lint-judge"
remote = "getsentry/skills"
scope = "report"
reportOn = "low"

[[skills.triggers]]
type = "pull_request"
Expand Down
Loading