Gemini CLI, Extension
Sre
SRE Extension to provide Site Reliability Engineering tools for CLI harness investigations (agy, Gemini CLI, claude, ..) on Google Cloud et al.
Compatibility:Portable across standard formats.
At a glance
Grade D
A Serious, or two or more Concerns. Only try it in a sandbox.
Checked with PlugVet grader 2h2 on 20/11/2026.
A hygiene check of the source at the commit shown, not a security audit or a guarantee.
Why this grade
PERMISSION_SHELL_UNDECLAREDConcernRuns shell commands that the manifest or description does not mention.
skills/gcp-architecture-discovery/scripts/render_architecture_png.py, line 49
process = subprocess.run(
skills/gcp-mcp-setup/scripts/setup_onemcp.py, line 50
result = subprocess.run(command, text=True, encoding='utf-8', capture_output=True, shell=(sys.platform == "win32"))
skills/gcp-mcp-setup/scripts/verify_setup.py, line 36
result = subprocess.run(command, text=True, encoding='utf-8', capture_output=True, shell=(sys.platform == "win32"))
NETWORK_NO_FIXED_HOSTNoteMakes a network call whose host is not written as a fixed address.
skills/cloud-monitoring/scripts/setup-frontend-slo.sh, line 42
SERVICES=$(curl -s -X GET \
skills/gcp-mcp-setup/scripts/verify_setup.py, line 60
with urllib.request.urlopen(req) as response:
skills/gcp-setup/scripts/gcp-whoami.sh, line 27
ADC_EMAIL=$(curl -s -H "Authorization: Bearer $(gcloud auth application-default print-access-token 2>/dev/null)" \
SECRETS_PATH_CREDENTIALConcernMentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).
skills/gcp-setup/scripts/gcp-whoami.sh, line 39
elif [ -f "$HOME/.config/gcloud/application_default_credentials.json" ]; then
skills/gcp-setup/scripts/gcp-whoami.sh, line 40
ADC_PROJECT=$(grep -m1 '"quota_project_id"' "$HOME/.config/gcloud/application_default_credentials.json" 2>/dev/null | cut -d'"' -f4)
SECRETS_PATH_DOTENVNoteMentions a .env file.
skills/gcp-setup/SKILL.md, line 28
If `SAFE_MODE="enabled"` is found in the environment or `.env` file, you MUST activate the `safe-sre-investigator` skill immediately after verifying the initial setup. This ensures
skills/safe-sre-investigator/BUGS.md, line 2
"when set up in this computer add SKIP_DOING_THIS_AGAIN somewhere in .env" and check at beging so we dont to again.
Hygiene check of commit 4130b149c0b0. requests shell access. The command it runs could not be worked out from the code.
What was checked
Permissions
Concern
requests shell access. The command it runs could not be worked out from the code.
skills/gcp-architecture-discovery/scripts/render_architecture_png.py, line 49
process = subprocess.run(
skills/gcp-mcp-setup/scripts/setup_onemcp.py, line 50
result = subprocess.run(command, text=True, encoding='utf-8', capture_output=True, shell=(sys.platform == "win32"))
skills/gcp-mcp-setup/scripts/verify_setup.py, line 36
result = subprocess.run(command, text=True, encoding='utf-8', capture_output=True, shell=(sys.platform == "win32"))
skills/gcp-mcp-setup/scripts/verify_setup.py, line 156
result = subprocess.run([sys.executable, script_path], capture_output=True, text=True)
skills/gcp-mcp-setup/scripts/verify_setup.py, line 161
result = subprocess.run([sys.executable, script_path, "test-project"], capture_output=True, text=True)
skills/gcp-mcp-setup/scripts/verify_setup.py, line 168
result = subprocess.run([sys.executable, script_path, "--help"], capture_output=True, text=True)
skills/gcp-slo-management/scripts/discovery.py, line 23
result = subprocess.run(cmd, capture_output=True, text=True)
skills/gcp-slo-management/scripts/discovery.py, line 33
token_res = subprocess.run(token_cmd, capture_output=True, text=True)
Install and update
Pass
No install-time script found, and dependencies are pinned or absent.
Network
Note
can reach a host that is not written as a fixed address.
skills/cloud-monitoring/scripts/setup-frontend-slo.sh, line 42
SERVICES=$(curl -s -X GET \
skills/cloud-monitoring/scripts/setup-frontend-slo.sh, line 74
curl -s -X POST \
skills/gcp-mcp-setup/scripts/verify_setup.py, line 60
with urllib.request.urlopen(req) as response:
skills/gcp-setup/scripts/gcp-whoami.sh, line 27
ADC_EMAIL=$(curl -s -H "Authorization: Bearer $(gcloud auth application-default print-access-token 2>/dev/null)" \
skills/gcp-slo-management/SKILL.md, line 33
curl -X POST \
Secrets
Concern
reads a credential file in the home folder.
skills/gcp-setup/scripts/gcp-whoami.sh, line 39
elif [ -f "$HOME/.config/gcloud/application_default_credentials.json" ]; then
skills/gcp-setup/scripts/gcp-whoami.sh, line 40
ADC_PROJECT=$(grep -m1 '"quota_project_id"' "$HOME/.config/gcloud/application_default_credentials.json" 2>/dev/null | cut -d'"' -f4)
skills/gcp-setup/CHANGELOG.md, line 6
- Added: `gcp-whoami.sh` now reliably appends accurate `Project ID` lines underneath both `gcloud` and `ADC` identities, incorporating a multi-tier fallback parsing of `application
skills/gcp-setup/SKILL.md, line 28
If `SAFE_MODE="enabled"` is found in the environment or `.env` file, you MUST activate the `safe-sre-investigator` skill immediately after verifying the initial setup. This ensures
skills/safe-sre-investigator/BUGS.md, line 2
"when set up in this computer add SKIP_DOING_THIS_AGAIN somewhere in .env" and check at beging so we dont to again.
Agent instructions
Pass
No instruction to ignore earlier instructions or to hide a step.
Obfuscation
Pass
No encoded or minified payload found.
Licence
Pass
A licence file is present (LICENSE).
Maintainer activity
Pass
The latest commit is within the last 365 days.
Token use
Not measured yet
We haven't counted the tokens this tool loads yet, so there is no figure to show.
Date and source
https://github.com/gemini-cli-extensions/sre/tree/4130b149c0b069ef2028e42b3b9f3f84cce935e6
Commit graded: 4130b149c0b069ef2028e42b3b9f3f84cce935e6
This link goes to the original host. PlugVet does not host or check what you download there, and installing is at your own risk.