github/gh-aw-firewall
GitHub Agentic Workflows Firewall
- Repository: https://github.com/github/gh-aw-firewall
- Commit:
a7a2bcad920d51a5968829a5859828212b50495f - Generated: 2026-11-14T00:10:33+00:00
- Scorer 0.1.0, scoring prov-4+qual-2+sec-2, report schema v2
Scores
| Pillar | Score | Signal | Confidence | Coverage |
|---|---|---|---|---|
| AI provenance (estimated AI-generated share) | 96 | explicit | 95% | 100% |
| Code quality | 74 | adequate | 90% | 100% |
| Security | 46 | poor | 85% | 100% |
AI provenance
Explicit AI attribution on 96% of first-party code added; an estimated 96% including velocity signals.
- Explicitly attributed share (lower bound): 96.4%
- Unattributed share in AI-consistent velocity bursts: 0.8%
- Estimated AI-generated share: 96.4%
| AI tool | Attributed commits | Lines added |
|---|---|---|
| GitHub Copilot coding agent | 2,383 | 554,713 |
| Claude Code | 203 | 118,829 |
- AI tooling configured in repo: claude-code, generic-agent
Current source lines last changed by AI-attributed commits: 174,850 of 179,024.
| Findings per 1,000 lines | AI-attributed | Other | Ratio |
|---|---|---|---|
| Quality | 1.132 | 0.479 | 2.36x |
| Security | 1.584 | 4.312 | 0.37x |
- Stylometry index: 0.43 per 1,000 lines (narrated steps 20, placeholders 0, assistant phrasing 0, emoji 114). Reported as a signal, not scored.
| Month | Lines added | Attributed to AI |
|---|---|---|
| 2025-12 | 9,159 | 100% |
| 2026-01 | 96,308 | 100% |
| 2026-02 | 13,236 | 98% |
| 2026-03 | 12,823 | 100% |
| 2026-04 | 14,909 | 99% |
| 2026-05 | 76,115 | 100% |
| 2026-06 | 50,879 | 99% |
| 2026-07 | 46,574 | 86% |
| 2026-08 | 79,434 | 100% |
| 2026-09 | 67,104 | 97% |
| 2026-10 | 149,793 | 96% |
| 2026-11 | 62,814 | 100% |
Evidence
- GitHub Copilot coding agent: trailer 'Co-authored-by: copilot' (a7a2bcad920d)
- GitHub Copilot coding agent: author Copilot (13df8d5dea42)
- GitHub Copilot coding agent: author Copilot (9e8c2e80b010)
- GitHub Copilot coding agent: author Copilot (13b097efb80a)
- GitHub Copilot coding agent: author Copilot (5b47bfd0f7c4)
- GitHub Copilot coding agent: author Copilot (1e9c6762080f)
- GitHub Copilot coding agent: author Copilot (f788a95e93a2)
- GitHub Copilot coding agent: author Copilot (0592ead4bdd6)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (e4dc5a476e1c)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (4fbb01825867)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (ef8519d2dd42)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (03dadabe3c25)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (3a3fca194f43)
- GitHub Copilot coding agent: trailer 'Co-authored-by: Copilot' (4e04a30fdd01)
- GitHub Copilot coding agent: author Copilot (9d7a20e0fc83)
AI provenance describes how code was produced; it is not a quality judgment. Unless attribution is explicit, these are signals consistent with AI generation, not proof. Squash/rebase workflows, imported code, and rewritten history reduce accuracy.
Code quality
Code quality is adequate: test-to-source ratio 1.36, CI runs tests; 28% of function code in complex or long functions; 8% duplicated lines. Weakest area: complexity.
| Area | Score |
|---|---|
| Complexity (lizard) | 43 |
| Duplication | 63 |
| Code smells (Opengrep) | 70 |
| Tests and CI | 100 |
| Documentation | 80 |
| Type safety | 97 |
| Process (Scorecard review/maintenance) | n/a |
- Functions: 9,757; mean complexity 4.2, 90th percentile 9; 437 complex and 86 long functions
- Duplicated lines: 7.6% of 135,349 compared
- Tests: 244,335 test lines for 179,024 source lines (ratio 1.36); CI test commands: npm test, npm run test, go test, jest; 98% of recent source commits also change tests
- Docs missing or thin: changelog
Most complex functions:
- checkRequest in containers/api-proxy/protocol-compatibility/assessment.js: complexity 172, 183 lines
- compile in scripts/model-information/compile.cjs: complexity 168, 440 lines
- execute in src/cloud-hypervisor/host-enclave-executor.ts: complexity 111, 336 lines
- collect in scripts/ci/collect-enclave-smoke-debug.cjs: complexity 100, 224 lines
- validateRecord in src/cloud-hypervisor/cleanup-identity.ts: complexity 100, 140 lines
Security
Security posture is poor: 14 known vulnerabilities in 9 dependencies; 27 high-severity SAST findings; 20 high-severity workflow issues. Weakest area: ci actions.
| Area | Score |
|---|---|
| Secrets (gitleaks, tree + history) | 44 |
| Static analysis (Opengrep) | 39 |
| Vulnerable dependencies (OSV) | 64 |
| CI workflows (zizmor) | 0 |
| Posture (policy, updates, signing, lockfiles) | 100 |
| OpenSSF Scorecard | n/a |
-
Known vulnerabilities: 14 in 9 packages (critical 0, high 0, medium 9, low 3)
-
Posture checks passed: security policy, dependency updates, signed commits; missing: lockfiles
Top findings
| Severity | Finding | Location |
|---|---|---|
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'run' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
.claude/skills/awf-debug-tools/scripts/common.py:360 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'run' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
.claude/skills/awf-debug-tools/scripts/common.py:368 |
| high | opengrep/dockerfile.security.missing-user-entrypoint: By not specifying a USER, a program in the container may run as 'root'. This is a security hazard. If an attacker can control a process running as root, they may have control over the container. Ensure that the last USER in a Dockerfile is a USER other than 'root'. [confidence medium] |
containers/agent/Dockerfile:310 |
| high | opengrep/dockerfile.security.missing-user: By not specifying a USER, a program in the container may run as 'root'. This is a security hazard. If an attacker can control a process running as root, they may have control over the container. Ensure that the last USER in a Dockerfile is a USER other than 'root'. [confidence medium] |
containers/dind-ubuntu/Dockerfile:34 |
| high | opengrep/dockerfile.security.last-user-is-root: The last user in the container is 'root'. This is a security hazard because if an attacker gains control of the container they will have root access. Switch back to another user after running commands as 'root'. [confidence medium] |
containers/enclave/Dockerfile:137 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'TimeoutExpired' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1146 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'TimeoutExpired' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1176 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'TimeoutExpired' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1181 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'Popen' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1398 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-tainted-env-args: Detected subprocess function 'Popen' with user controlled data. A malicious actor could leverage this to perform command injection. You may consider using 'shlex.quote()'. [confidence medium] |
containers/enclave/agent-entrypoint.py:1399 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'CompletedProcess' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1426 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-tainted-env-args: Detected subprocess function 'CompletedProcess' with user controlled data. A malicious actor could leverage this to perform command injection. You may consider using 'shlex.quote()'. [confidence medium] |
containers/enclave/agent-entrypoint.py:1426 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-audit: Detected subprocess function 'CompletedProcess' without a static string. If this data can be controlled by a malicious actor, it may be an instance of command injection. Audit the use of this call to ensure it is not controllable by an external resource. You may consider using 'shlex.escape()'. [confidence low] |
containers/enclave/agent-entrypoint.py:1444 |
| high | opengrep/python.lang.security.audit.dangerous-subprocess-use-tainted-env-args: Detected subprocess function 'CompletedProcess' with user controlled data. A malicious actor could leverage this to perform command injection. You may consider using 'shlex.quote()'. [confidence medium] |
containers/enclave/agent-entrypoint.py:1445 |
| high | opengrep/python.lang.security.audit.dangerous-os-exec-audit: Found dynamic content when spawning a process. This is dangerous if external data can reach this function call because it allows a malicious actor to execute commands. Ensure no external data reaches here. [confidence low] |
guest/cloud-hypervisor/verify-guest-engine.py:128 |
| high | opengrep/go.lang.security.audit.dangerous-syscall-exec: Detected non-static command inside Exec. Audit the input to 'syscall.Exec'. If unverified user data can reach this call site, this is a code injection vulnerability. A malicious actor can inject a malicious script to execute arbitrary code. [confidence low] |
guest/microvm-supervisor/runtime_linux.go:72 |
| high | opengrep/go.lang.security.audit.dangerous-exec-command: Detected non-static command inside Command. Audit the input to 'exec.Command'. If unverified user data can reach this call site, this is a code injection vulnerability. A malicious actor can inject a malicious script to execute arbitrary code. [confidence low] |
guest/microvm-supervisor/runtime_linux.go:344 |
| high | opengrep/go.lang.security.audit.dangerous-exec-command: Detected non-static command inside Command. Audit the input to 'exec.Command'. If unverified user data can reach this call site, this is a code injection vulnerability. A malicious actor can inject a malicious script to execute arbitrary code. [confidence low] |
guest/microvm-supervisor/runtime_linux.go:542 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument cmd. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
scripts/ci/benchmark-performance.ts:60 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument name. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
scripts/ci/benchmark-performance.ts:194 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument command. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
scripts/ci/enclave-smoke/commands.js:9 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument input. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
scripts/ci/nvx-matched-runtime-adapter.ts:504 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument input. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
scripts/ci/nvx-matched-runtime-adapter.ts:561 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument command. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
src/nvx/launch-executor.ts:112 |
| high | opengrep/javascript.lang.security.detect-child-process: Detected calls to child_process from a function argument request. This could lead to a command injection if the input is user controllable. Try to avoid calls to child_process, and if it is needed ensure user input is correctly sanitized or sandboxed. [confidence low] |
src/nvx/one-shot-process.ts:28 |
Static signals only: the scorer never runs the project's code. Secret findings are pattern matches that were not verified, and secret values are never stored.
Distribution signals
- binary-release-asset: release v0.29.13 ships awf.tgz (0 downloads)
- binary-release-asset: release v0.29.13 ships cloud-hypervisor-enclave-rootfs-x86_64.tar.gz (0 downloads)
- binary-release-asset: release v0.29.13 ships cloud-hypervisor-test-x86_64.tar.gz (5 downloads)
- binary-release-asset: release v0.29.13 ships enclave-rootfs.ext4.gz (3 downloads)
- binary-release-asset: release v0.29.13 ships nvx-test-x86_64.tar.gz (0 downloads)
Signals consistent with a repository that exists to distribute a download. They are heuristics, not a verdict; release files were not downloaded or inspected.
Repository profile
- Source lines: 179,024; test lines: 244,335
| Language | Lines |
|---|---|
| TypeScript | 273,235 |
| JavaScript | 117,646 |
| Shell | 14,081 |
| Go | 12,317 |
| Python | 5,131 |
| Rust | 486 |
| C | 463 |