MCP, MCP server
checkexploit
CheckExploit — proves whether a scanner-flagged CVE is actually reachable in a repo, runs real exploits in an isolated sandbox to confirm, then patches and verifies fixes.
Compatibility: Not stated.
Grade C
One Medium finding. Read the report before installing.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
SECRETS_PATH_CREDENTIAL
Why this grade
- SECRETS_PATH_CREDENTIAL Medium
Mentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).
agent/analyzer/gen_context.py, line 653
"*.pfx", "id_rsa*", ".aws", ".ssh", ".npmrc",
Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.
Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.
What was checked
Plain notes from the hygiene checks.
Permissions
Concern
can delete files.
PERMISSION_DELETE_FILES
agent/build_image.py, line 19
return subprocess.check_output(
agent/build_image.py, line 33
r = subprocess.run(
agent/build_image.py, line 227
subprocess.run(
Show more
agent/build_image.py, line 231
shutil.rmtree(ctx, ignore_errors=True)
agent/orchestrate.py, line 38
subprocess.run(["git", "clone", "--quiet", repo_arg, str(clone_dir)], check=True)
agent/scan.py, line 25
subprocess.run(
agent/scan.py, line 39
subprocess.run(
agent/mcp/local_sandbox_server.py, line 91
proc = subprocess.run(
agent/mcp/local_sandbox_server.py, line 382
proc = subprocess.run(
agent/mcp/local_sandbox_server.py, line 387
shutil.rmtree(clone_root, ignore_errors=True)
agent/mcp/local_sandbox_server.py, line 473
shutil.rmtree(tmp_dir, ignore_errors=True)
Install and update
Pass
No install-time script found, and dependencies are pinned or absent.
Network
Concern
can reach a host that is not written as a fixed address.
NETWORK_NO_FIXED_HOST
agent/analyzer/reach.py, line 132
with urllib.request.urlopen(CVEORG_QUERY_URL + cve_id, timeout=15) as resp:
agent/analyzer/reach.py, line 149
with urllib.request.urlopen(OSV_QUERY_URL + cve_id, timeout=15) as resp:
agent/analyzer/reach.py, line 755
with urllib.request.urlopen(req, timeout=15) as resp:
Show more
agent/analyzer/reach.py, line 773
with urllib.request.urlopen(
agent/mcp/cve_feed_server.py, line 104
with urllib.request.urlopen(req, timeout=30) as resp:
scripts/mcp_client.py, line 15
with urllib.request.urlopen(req, timeout=300) as resp:
README.md, line 124
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8081/health').status)"
README.md, line 125
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8091/health').status)"
README.md, line 159
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8081/health').status)"
README.md, line 160
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8091/health').status)"
README.md, line 203
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8081/health').status)"
README.md, line 204
python3 -c "import urllib.request; print(urllib.request.urlopen('http://127.0.0.1:8091/health').status)"
agent/skills/reproducer/SKILL.md, line 27
5. `sandbox_exec` (health check): `python3 -c "import urllib.request; urllib.request.urlopen('http://127.0.0.1:PORT/health')"`
Secrets
Concern
reads a credential or environment file.
SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
agent/build_image.py, line 221
ignore=shutil.ignore_patterns(".env*", "__pycache__", "node_modules"))
agent/analyzer/gen_context.py, line 652
"\n".join((".git", ".env", ".env.*", "*.pem", "*.key", "*.p12",
agent/analyzer/gen_context.py, line 653
"*.pfx", "id_rsa*", ".aws", ".ssh", ".npmrc",
Show more
agent/mcp/local_sandbox_server.py, line 435
# dot-prefixed names like '..env' while rejecting any
.env.example, line 1
# Copy to .env and fill in. Never commit .env.
.gitignore, line 5
# Environment / secrets
.opencode/agents/checkexploit.md, line 44
- **Secrets never reach the sandbox.** Do not mount `.env` or credentials.
AGENTS.md, line 28
- **Never display secrets**: refer to API keys by name only, never print `.env` or tokens.
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.
Note
Hygiene check of commit d9017d46d694. can delete files. can reach a host that is not written as a fixed address. reads a credential or environment file.
Date and source
Date checked:
github.com
Commit graded: d9017d46d694e9510e273d92027c1e2c860d6d95
This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.